Mini Review

Cellular & Molecular Immunology (2015) 12, 525–532; doi:10.1038/cmi.2015.12; published online 16 March 2015

Regulatory T cells turn pathogenic

Jitao Guo and Xuyu Zhou

Key Laboratory of Pathogenic Microbiology and Immunology, Institute of Microbiology, Chinese Academy of Sciences (CASPMI), Beijing, China

Correspondence: Dr XY Zhou, CAS Key Laboratory of Pathogenic Microbiology and Immunology, Institute of Microbiology, Chinese Academy of Sciences, No.1 Beichen West Rd, Chaoyang District, Beijing 100101, China. E-mail:

Received 28 October 2014; Revised 21 January 2015; Accepted 23 January 2015
Advance online publication 16 March 2015



Foxp3+ regulatory T (Treg) cells are considered a sub-lineage of CD4+ T cells that are protective against autoimmunity due to their essential roles in maintaining immune homeostasis and self-tolerance. However, Treg cells are unstable in vivo in terms of lineage specialization and suppressive function. These unstable Treg cells play roles in the pathogenesis of diseases, which cause safety concerns regarding human Treg cell therapy. In this review, we highlight recent findings that demonstrate the pathogenic conversion of Treg cells in different disease models.


autoimmune; Foxp3; instability; pathogenic; Treg cell



The T cell-mediated immune response is tightly controlled, ensuring the protection of the body while limiting self-harm, which depends on several mechanisms of immune tolerance, including clonal deletion, clonal anergy, immunological ignorance and suppressive T cells.1 Several types of suppressive T cells have been reported, such as IL-10-producing CD4+ Tr1 cells,2 TGF-β-producing CD4+ Th3 cells,3 CD4+CD25+Foxp3+ T cells,4,5 CD8+CD28 T cells,6 and Qa-1-restricted CD8+ T cells.7 Among these suppressive T cells, CD4+CD25+Foxp3+ regulatory T (Treg) cells are the most physiologically relevant due to their broad and indispensable roles in vivo. Hence, we focus on Treg cells in this review.

As a sub-lineage of CD4+ T cells, Treg cells develop in the thymus (tTreg) or are induced in the periphery from naive T cells (pTreg), which constitute approximately 10% of the peripheral CD4+ T cells in humans and mice.4,8 Treg cells generated ex vivo, such as those induced with TGF-β, are termed ‘in vitro-induced Treg cells (iTreg)’, which can also suppress autoimmune disease in vivo.9,10 Treg cells specifically express the transcription factor Foxp3 (forkhead box protein P3) and highly express surface and intracellular markers, including CD25, GITR and CTLA-4, which define the phenotypes and functions of Treg cells.11,12,13 In humans, CD4+ CD45RA+FoxP3lo and CD45RAFoxP3hi cells are considered bona fide regulatory T cells because activated T cells can upregulate FOXP3 expression without acquiring a regulatory phenotype.14 The suppressive functions of Treg cells play crucial roles in the maintenance of immune homeostasis and self-tolerance, as demonstrated by scrufy mice and IPEX (immune dysregulation, polyendocrinopathy, enteropathy, X-linked syndrome) patients harboring Foxp3 gene mutations and develop severe autoimmune diseases because of the paucity or dysfunction of Treg cells.15,16 Treg cells do not have a unified mechanism of immune suppression. In contrast, they utilize diverse suppressive mechanisms depending on microenvironments and target cells, including inhibitory cytokines (e.g., IL-10, IL-35 and TGF-β), cytolysis, metabolic disruption or the modulation of dendritic cell function. Based on these striking immune suppressive properties of Treg cells, several clinical trials utilizing them to treat autoimmune diseases are currently underway and show promising potential.17

Recently, increasing evidence has indicated that the lineage specialization and suppressive function of Treg cells are unstable in vivo, as demonstrated by the loss of Foxp3 expression and acquisition of Th cell-like functions, which underscore the safety concerns with regard to human Treg cell therapy. Although the instability of Treg cells may be physiologically required for immune homeostasis (e.g., transient downregulation of Treg cell function favors an effective immune response against pathogen invasion), it should be noted that Treg cell instability can be pathogenic. Indeed, several studies report on the pathogenic cells derived from Treg cells either in mice or in humans. In this review, we highlight the studies that have demonstrated the pathogenic conversion of Treg cells in different disease models.


Functions of Treg cells and health

Before discussing the pathogenic conversion of Treg cells, it is important to understand the normal functions of Treg cells. Consistent with the severe autoimmune diseases in scurfy mice and IPEX patients, the ablation of Treg cells through germ-line deletion of the Foxp3 gene, conditional deletion of Foxp3 in mature Treg cells, or diphtheria toxin receptor-mediated depletion of Treg cells, leads to severe autoimmunity in mice.18,19,20 In contrast, transferred Treg cells, Foxp3 transgene expression, or bone marrow reconstitution can rescue the scurfy mice,15,18,21 indicating that Treg cells are continuously needed throughout life. Moreover, as a subset of lymphocytes, Treg cells circulate throughout the body and generate locally. Therefore, they are located in different anatomical sites, such as blood, lymphoid organs and diverse tissues (where they stay in different developmental stages and functional statuses), to safeguard the body against autoimmunity. The necessity of Treg cells for the body primarily depends on their immune suppressive function, which can target T cells, dendritic cells and other immune cells. Treg cells help to maintain immune homeostasis and self-tolerance and cannot be compensated for by other suppressive mechanisms when Treg cells are absent in vivo. Collectively, Treg cells contribute to health as a dominant immune regulator.

The detailed roles of Treg cells in immune regulation have been found gradually in recent years. Evidence supports the notion that Treg cells can distinguish between different types of immune responses when they execute their immunosuppressive function, which is consistent with the fact that Treg cells represent heterogeneous populations sharing certain common features, such as high expression of Foxp3 and CD25, while maintaining certain unique features in individual subpopulations. It appears that Treg cells can utilize the lineage-determining transcription factors of other cell lineages to equip themselves for their immune regulation targeting different types of immune response in various anatomical sites. For example, T-bet-expressing Treg cells selectively monitor the Th1 immune response by up-regulating CXCR3.22 Similarly, the Th2 transcription factor IRF4 is needed for Treg cells to inhibit the Th2-mediated autoimmune response,23 and other transcription factors such as STAT3 and Bcl-6 are necessary to inhibit Th17-mediated spontaneous colitis and aberrant germinal center responses, respectively.24,25,26 Moreover, Treg cells express tissue-associated transcriptions factor to perform different functions in tissues, such as PPAR-γ, the ‘master regulator’ of adipocyte differentiation and function.27 PPAR-γ-expressing Treg cells selectively accumulate in the visceral adipose tissue of aged mice and acquire adipocyte-like functions, such as taking up lipids.28 In addition, a recent study showed that a subset of Treg cells could infiltrate injured skeletal muscle and contribute to muscle repair. These muscle-resident Treg cells produce the growth factor Amphiregulin, which aids in the colony-forming efficiency of muscle satellite cells in vitro and improves muscle repair in vivo.29 These studies portray a scenario in which Treg cells are in charge of the immune balance throughout the entire body. To adapt to actual needs, Treg cells continue their differentiation, accompanied by Foxp3 expression, in diverse microenvironments. The result is that Treg cells obtain additional functions that control diverse immune and other physiologic responses.

Additional functions for Treg cells will be discovered in the future, especially their roles in various tissues. Despite their low frequency among total Treg cells, certain tissue-resident Treg cells play crucial roles in their resident microenvironments, which are more relevant to local-tissue disorders than total Treg cells.


Treg cells are potentially pathogenic

As a level of immune control, Treg cells are irreplaceable, which highlights the risk to the host of Treg cells not functioning properly. The scurfy/IPEX disease and plenty of experimental impairments of Treg cell function prove this scenario. However, an overly strong suppressive effect of Treg cells may hinder effective immune responses to pathogens. For example, the priming of T cells in draining lymph nodes is delayed due to expanded Mycobacterium tuberculosis (Mtb)-specific Treg cells during Mtb infection, and T cell migration to infected lungs is also impeded, resulting in deficient clearance of Mtb.30,31 In contrast, the ablation of Treg cells contributes to Mtb clearance.32 Moreover, Treg cells can accumulate at tumor sites to inhibit anti-tumor immune responses and facilitate tumor immune evasion.33 Aside from improper suppressive functions promoting disease progression indirectly, Treg cells have the potential to promote diseases directly by conversion into pathogenic cells, especially in autoimmune diseases. Some unique characteristics make Treg cells potentially pathogenic, including their linage instability, self-skewed T-cell receptor (TCR) repertoire, and atypical functions. These characteristics are discussed below.

Instability of the Treg cell lineage. Foxp3 is indispensable for the development and function of Treg cells. However, increasing evidence indicates that Treg cells can lose Foxp3 expression to generate so-called ‘exTreg’ cells in vivo, which are associated with external inputs, such as IL-2 depletion, strong TCR engagement with autoantigens, and pro-inflammatory cytokines.34,35,36 The TCR repertoire of exTreg cells substantially overlaps with that of both Treg cells and conventional T cells, which indicates that exTreg cells are heterogeneous and are derived in part from established Treg cells that have lost Foxp3 expression and in part from Foxp3 conventional T cells that expressed Foxp3 transiently.34,37

Foxp3 expression in Treg cells is maintained intrinsically by a positive feedback loop comprising the CNS2 region in the Foxp3 gene locus, the Cbfβ-Runx1 transcription factor, Foxp3 itself, and other factors if any, where the CNS2, Cbfβ-Runx1 and Foxp3 binds to each other to form a transcription complex.19,38,39,40,41 Treg cells lacking CNS2, Cbfβ, or Runx1 cannot maintain stable Foxp3 expression.38,39 Similarly, a recent study further highlighting the importance of CNS2 demonstrates that Treg cells are stabilized by the IL-2/STAT5 pathway while destabilized by the IL-4/STAT6 and IL-6/STAT3 pathways through the competitive occupation of CNS2 between STATs. In the absence of CNS2, Treg cells cannot maintain heritable Foxp3 expression either under IL-2-limiting conditions or in the presence of inflammatory cytokines.42 Another relevant study shows that the NFAT-mediated looping between CNS2 and the Foxp3 promoter is critical for Foxp3 expression in activated Treg cells.43 These studies indicate that the CNS2-mediated feedback loop is critical for the maintenance of Treg cell lineage stability.

Attenuating the DNA binding activity of Foxp3 potentially breaks the CNS2–Cbfβ–Runx1–Foxp3 feedback loop, resulting in Treg cell lineage instability. We developed a luciferase-based reporter system (FOXP3Luc) to monitor the DNA binding activity of Foxp3. Using this reporter for an unbiased screening, we identified that MAPKK kinase COT/Tpl2, and its target MEK1, showed inhibitory effects on the readout of the FOXP3Luc reporter. The inhibition of either COT/Tpl2 or MEK1 favors the stable Foxp3 expression in cultured Treg cells. Importantly, constitutive activation of MEK1 destabilizes Treg cells in vivo.44 This finding is consistent with our previous study indicating the role of TCR overstimulation by autoantigens in Treg cell instability; BDC2.5 TCR+ Treg cells or MOG38–49-specific Treg cells, which recognize autoantigens of the pancreas and myelin oligodendrocyte glycoprotein in the central nervous system, respectively, are significantly unstable in inflamed sites.34,45 Recent studies report that PTEN is essential for maintaining Treg cell stability by limiting phosphoinositide 3-kinase signaling,46,47 which also supports the notion that TCR overstimulation destabilizes Treg cells. In addition, the CDK2-mediated Foxp3 degradation and PIM1-mediated down-regulation of DNA binding activity of human FOXP3 are potentially involved in Treg cell instability.48,49

With the loss of Foxp3 expression, Treg cells usually lose their suppressive function and obtain partial phenotypes and functions of effector Th cells, such as the production of IFN-γ and IL-17. Given the crucial roles of Treg cells in immune balance, the instability of Treg cells, especially in local tissues, such as pancreas islets, disturbs the immune balance and unleashes a local pathogenic T-cell response. In addition, the effector Th cell-like functions of ‘exTreg’ cells also directly contribute to the pathogenic immune response in local tissues.

Self-skewed TCR repertoire. The TCR endows T cells with high specificity to recognize diverse antigens, and TCR stimulation by antigens leads to T-cell activation in the presence of co-stimulatory signals in the periphery. In the central tolerance process, thymocytes harboring autoreactive TCRs are subjected to apoptosis during negative selection in the thymus, whereas those with proper TCR specificity and affinity survive and mature into T cells, including Treg cells and naive T cells. However, there are differences between the TCR properties of Treg cells and naive T cells, as demonstrated by evidence from TCR repertoire sequencing, TCR transgenic mice, and cell lines, which suggest that high-affinity TCRs below the threshold required for negative selection favor the development of Treg cells in the thymus, resulting in a TCR repertoire that is biased toward self-antigens in Treg cells.50,51,52,53 Although it remains unclear how such a self-skewed TCR repertoire contributes to the function of Treg cells in the maintenance of immune homeostasis and self-tolerance, it should be noted that these self-reactive TCRs are potential threats for the host if the Treg cells lose Foxp3 expression and turn into effector-like Th cells. ‘exTreg’ cells may be more harmful than conventional T cells because they are prone to activation by autoantigens due to their autoreactive TCRs and effector Th cell-like functions.

Atypical function. Treg cells can be reshaped by microenvironments where they reside in the presence of Foxp3 expression, which may account for the ‘plasticity’ or ‘heterogeneous’ property of this cell lineage. As mentioned above, Treg cells obtain adipocyte-like functions in visceral adipose tissue and produce Amphiregulin in injured muscle. Consistent with this notion, it is possible that Treg cells subjected to improper “education” from pathogenic environments, such as tumor microenvironments, can acquire detrimental functions to promote diseases. For example, Facciabene et al.54 showed that tumor-recruited Treg cells expressed vascular endothelial growth factor A, which directly contributed to tumor angiogenesis. Similarly, Tan et al.55 reported that tumor-infiltrating Treg cells expressed RANKL, which stimulated the metastatic progression of RANK-expressing breast/mammary carcinoma cells. Moreover, Blatner et al.56 described a Treg cell subset marked by the expression of RORγt in human colon cancer and a mouse model of hereditary polyposis. These RORγt+Foxp3+ T cells retained suppressive function in vitro but were impaired in their anti-inflammatory properties in vivo. Their pathogenic roles in hereditary polyposis were also indirectly demonstrated in that the ablation of the RORγt gene in Foxp3+ cells rescued polyp-prone mice from polyposis.


Pathogenic exTreg cells and diseases

Consistent with the potential pathogenicity of Treg cells discussed above, many studies have proved that Treg cells indeed turn pathogenic in several disease models. In most cases reported to date, the pathogenic conversion of Treg cells is associated with an instability of Foxp3 expression. According to Koch's postulates,57 bacteria that are sufficient to introduce disease in a healthy organism are recognized as pathogenic bacteria. Similarly, to define the pathogenicity of exTreg cells, it should be clear whether these cells are able to transfer diseases when they are introduced into a healthy host. Because lymphocytes can be easily transferred into a new host through intravenous injection, most studies are based on adoptive transfer experiments in lymphodeficient mice (Table 1).

Foxp3-GFP KI reporter mice harbor GFP-labeled Treg cells expressing a chimeric GFP-Foxp3 fusion protein,21 which helps to obtain Treg cells with high-purity. Using this mice strain, Zheng et al.58,59 showed that nTreg cells are converted to Th17 cells by IL-6 treatment in vitro. To study the Treg cell stability in vivo, Duarte et al.60 purified CD4+GFP+ Treg cells from Foxp3-GFP KI reporter mice and transferred them into syngenic Rag2−/− recipient mice, which are deficient for T and B cells. After 4 weeks, they found that approximately half of the transferred CD4+GFP+ Treg cells lost Foxp3 expression. To study the function of these converted Treg cells, they purified Treg-derived CD4+Foxp3 T cells (Foxp3+right arrowFoxp3) from the Rag2−/−recipient mice injected with CD4+GFP+ Treg cells and subsequently transferred these resultant Foxp3 T cells into new Rag2−/− recipient mice. These Treg-derived CD4+Foxp3 T cells acquired the ability to produce IL-4, IFN-γ and TNF-α upon this secondary transfer. Importantly, Rag2−/− mice that received Treg-derived CD4+Foxp3 T cells developed lung and liver inflammation, as did the mice that were reconstituted with conventional T cells. In contrast, the transfer of Treg cells purified from Foxp3-GFP KI mice or recovered from Rag2−/− mice did not lead to inflammation. Similar results were obtained from TCR-β−/−recipient mice, indicating that B cells are not responsible for the instability of Treg cells in vivo. This study provides direct evidence for the pathogenicity of exTreg cells, but the transferred CD4+Foxp3 T cells do not represent the naturally generated exTreg cells in vivo because their generation is primarily due to insufficient IL-2 in recipient mice.60

To trace Treg cells in vivo, we generated another Foxp3 reporter mouse strain that carries Foxp3 BAC-driven GFP-Cre. In this strain, the GFP signal indicates transcription of the Foxp3 gene with high fidelity.61 These mice were crossed to Rosa26-loxp-stop-loxp-YFP mice62 to generate Foxp3-GFP-Cre×R26-YFP mice in which the GFP-Cre+ cells were permanently marked by YFP due to the removal of the floxed-stop cassette by the GFP-Cre activity. Thus, Foxp3-GFP-Cre×R26-YFP mice provide a valuable model to monitor Foxp3 expression in traceable Treg cells at the single-cell level. Using this model, we found that considerable exTreg cells existed in the periphery in vivo under both homeostatic and autoimmune conditions. These exTreg cells resembled activated/memory T cells in respect to their phenotypes and their ability to produce inflammatory cytokines. To further investigate the function of these exTreg cells, we sorted exFoxp3 cells, conventional T cells, and Treg cells from Foxp3-GFP-Cre×R26-YFP BDC2.5 mice (BDC2.5 TCR-transgenic CD4+ T cells recognize a pancreatic islet autoantigen), and transferred each population into NOD Rag2−/− mice after expansion in vitro. We observed that mice receiving Treg cells had normal blood glucose concentrations, whereas those receiving exFoxp3 cells and conventional T cells rapidly displayed islet destruction and developed diabetes.34 Our study describes exTreg cells physiologically generated in vivo and proves that such natural exTreg cells are pathogenic. The exFoxp3 cell expressing BDC2.5 TCR specifically damages pancreatic islet, consistent with the concerns about the skewed self-reactive TCR repertoire of exTreg cells.

Similarly, the pathogenicity of exTreg cells is also evident in an experimental autoimmune encephalitis (EAE) model.45 In this study, Bailey-Bucktrout et al.45 induced EAE disease in Foxp3-GFP-Cre×R26-RFP mice using MOG35–55 peptides and traced the MOG35–55-specific T cell population by I-Ab-MOG38–49 tetramer staining. They found that MOG35–55-specific Treg cells were more unstable than polyclonal Treg cells isolated from the same inflammatory sites during the EAE induction phase, and exFoxp3 cells produced IFN-γ at a level comparable to conventional T cells when stimulated by MOG35–55 peptides. Further, conventional T cells, exFoxp3 cells and Treg cells from MOG35–55-CFA-immunized mice were compared for their ability to induce EAE disease upon transfer into EAE-inducing lymphodeficient recipients. The results showed that transferred exFoxp3 cells and conventional T cells, but not Treg cells, induced EAE with comparable incidence and severity. Despite the fact that the transferred exTreg cells in this and the above study are naturally generated in vivo, it is not clear whether the in vitro expansion procedure for adoptive transfer affects the quality of individual T cell populations.

Komatsu et al.63 reported more precise transfer experiments of exTreg cells without using in vitro expansion procedures. In this study, transferred CD25loFoxp3+ T cells more easily lost Foxp3 expression and converted to Th17 cells when compared with CD25hiFoxp3+ T cells in a collagen-induced arthritis model. The Treg cell-derived Th17 cells directly induced osteoclastogenesis in vitro by producing high levels of RANKL. In contrast, naive CD4+ T cell-derived Th17 cells alone did not induce osteoclastogenesis.64 To study the pathogenicity of unstable Treg cells in arthritis, CD25loFoxp3+, CD25hiFoxp3+, total Foxp3 and effector memory CD44hiFoxp3CD4+ T cells from collagen-immunized DBA/1 Foxp3hCD2 mice were purified and labeled by CFSE and transferred into immunized DBA/1 Foxp3hCD2mice 1 day before secondary immunization. The results demonstrated that transferred CD25loFoxp3+ T cells were more effective at exacerbating arthritic symptoms than total Foxp3 or CD44hiFoxp3CD4+ T cells, whereas CD25hiFoxp3+ cells alleviated arthritic symptoms. In contrast, OVA-specific CD25loFoxp3+CD4+ T cells did not exacerbate arthritis symptoms, indicating that the pathogenicity of exTreg cells in the collagen-induced arthritis model depended on their TCR specificity to type II collagen.

Human Treg cells are more likely easy to convert to Th17 cells by IL-1β and IL-2 treatment in vitro.65,66 Bovenschen et al.67 reported that Treg cells from psoriasis patients showed reduced suppressive function and an increased ability to produce IL-17A when compared with healthy controls. Ex vivo-stimulated Treg cells from psoriasis patients gradually obtained a high level of RORγt expression and lost Foxp3 expression, which suggested that Treg cell converted to Th17 cells through a Foxp3/RORγt-double-positive stage. Importantly, IL-17A-producing Foxp3+CD4+ T cells accumulated in the dermis of the lesional skin of severe psoriatic patients, which iindicated a pathogenic role of unstable Treg cells in human psoriatic disease.

Some studies on function-modified Treg cells also provide evidence for the potential pathogenicity of Treg cells. For instance, Takahashi et al.68 reported that SOCS1 (suppressor of cytokine signaling 1)-deficient Treg cells produced high levels of IFN-γ and rapidly lost Foxp3 when transferred into Rag2−/− mice. Importantly, Rag2−/−mice injected with Socs1−/− Treg cells developed extremely severe colitis. Despite the lack of direct evidence from adoptive transfer experiments, several reports have demonstrated a deficiency of key genes in Treg cells, such as smad2/3, TRAF6 and Foxo1, promoted Treg cells to produce IL-4 or IFN-γ, which support the pathogenic convention of Treg cells.69,70,71 Moreover, attenuated Foxp3 expression in Foxp3-IRES-luciferase-IRES-eGFP (FILIG) mice leads to impaired suppressive function of Treg cells and converts Treg cells into IL-4-secreting Th2-like effector T cells, which leads to Th2-mediated autoimmune diseases.72 Similarly, CNS2-deficient Treg cells readily lose Foxp3 expression and convert to IFN-γ- and IL-17-producing effector T cells under inflammatory conditions, which contributes to the autoimmune disease in CNS2-deficient mice.42,43

Taken together, these studies based on adoptive transfer experiments directly prove the pathogenicity of exTreg cells in vivo. It appears that both self-reactive TCR and effector Th cell-like functions are required for the pathogenic role of exTreg cells. It is possible that autoreactivity directs the activation and accumulation of exTreg cells in target tissues, and effector T helper cell-like functions of exTreg cells promote pathogenic immune responses.


ExTreg cells participate in a feed-forward loop in the pathogenesis of autoimmune diseases

Autoimmune diseases occur when self-antigens are targeted by the adaptive immune response. These self-antigens cannot be cleared, which leads to sustained autoimmune responses and tissue damage, such as pancreatic β-cell destruction in type 1 diabetes, joint inflammation in rheumatoid arthritis, and central nervous system tissue damage in human multiple sclerosis. The damaged tissues are usually infiltrated with diverse stromal and immune cells, pro-inflammatory cytokines, chemokines and exposed self-antigens, which constitute complex pathological microenvironments where the immune balance is skewed toward inflammation. Based on the pathogenicity of exTreg cells in autoimmune diseases, we propose a feed-forward model to understand the pathogenesis of autoimmune diseases. First, the exposed self-antigens and pro-inflammatory cytokines in damaged tissues continually activate effector Th cell-mediated immune responses, which in turn attack the self-tissues, resulting in more self-antigen exposure and production of pro-inflammatory cytokines in situ. Second, self-antigens and proinflammatory cytokines in damaged tissues promote the instability of local Treg cells, and unstable Treg cells turn pathogenic for the self-tissues, resulting in more severe damage. Third, the generation of exTreg cells favors the activation and expansion of autoreactive effector Th cells, on the other hand, activated effector Th cells promote the generation and effector Th-like function of exTreg cells by Th-producing cytokines (e.g., IL-4 and IL-642). In addition, it is possible that activated Th cells become more resistant to the suppression of Treg cells. Fourth, these feed-forward loops promote each other and constitute a higher-level feed-forward loop to drive the development of autoimmune diseases (Figure 1).

Figure 1.
Figure 1 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact or the author

The multilayer feed-forward loop model of autoimmune diseases. (1) Exposed autoantigens and pro-inflammatory cytokines in damaged tissues activate effector Th cells, which in turn exacerbate self-tissue damage. (2) Exposed self-antigens and pro-inflammatory cytokines in damaged tissues promote the generation of exTreg cells, which are pathogenic for the self-tissues. (3) exTreg cells favor the activation and expansion of autoreactive effector Th cells, which may further induce the generation and effector Th-like functions of exTreg cells through Th-producing cytokines. Together, these feed-forward loops constitute a higher-level feed-forward loop (4) that drives the development of autoimmune diseases forward. Th, T helper; Treg, regulatory T.

Full figure and legend (65K)


Generation of exTreg cells favors anti-tumor immunity

It is generally accepted that tumors recruit or induce Treg cells to facilitate their evasion of immune system attack. Both minimizing the existence and damping of the tumor-infiltrated Treg cell function are considered promising approaches to treating cancers. One paradox for such strategies is that drastic reduction of Treg cell immunoregulation may lead to autoimmunity. Thus, to elicit anti-tumor immunity without causing autoimmune disorders, modifying functions or stability of Treg cells is more optimal than systemically deleting Treg cells. To date, agonistic antibodies targeting CD25, GITR, CTLA-4, OX40 and FR4 have been reported to be useful for inhibiting the suppressive function of Treg cells and beneficial for anti-tumor immunity.73,74,75,76,77 Apart from antibodies blocking the key functional molecules of Treg cells, chemical drugs that selectively impede Treg cell function or stability are also attractive. C646, a small-molecule nonpeptidic inhibitor of p300 (p300i), destabilizes Foxp3 expression and impairs nTreg suppressive function; importantly, C646 treatment boosts anti-tumor immunity.78 Given that strong COT/Tpl2-MEK1 and PI(3)K signaling destabilizes Treg cells,44,46,47 approaches that specifically augment these signaling pathways should be helpful for inhibiting the immunoregulation of Treg cells and favoring anti-tumor immunity.



Although adoptive transfer experiments provide evidence for the pathogenicity of exTreg cells, it should be noted that lymphodeficient recipient mice and superabundant transferred exTreg cells may artificially exaggerate their pathogenicity. Nevertheless, studies using different disease models clearly show that Treg cells can become pathogenic. Such pathogenic conversion of Treg cells usually occurs in pathological environments, which implies a feedback mechanism between pathogenic conversion of Treg cells and diseases. Importantly, some findings provide valuable indications for solving this problem. For example, IL-2/IL-2 mAb complex and all-trans retinoic acid not only promote the expansion of Treg cells but also help to sustain the stable Foxp3 expression in Treg cells.45,79,80 Finally, pathogenic conversion of Treg cells is closely related to diseases caused by local tissue disorders so it is important to study Treg cells that reside in local pathological tissues rather than total Treg cells. More precise studies are needed to illustrate the roles of pathogenic conversion of Treg cells in diseases, which will aid in the development of medicines and therapies that target such pathogenically converted Treg cells.



  1. Bluestone JA. Mechanisms of tolerance. Immunol Rev 2011; 241: 5–19. | Article | PubMed | ISI |
  2. Roncarolo MG, Bacchetta R, Bordignon C, Narula S, Levings MK. Type 1 T regulatory cells. Immunol Rev 2001; 182: 68–79. | Article | PubMed | ISI | CAS |
  3. Weiner HL. Induction and mechanism of action of transforming growth factor-beta-secreting Th3 regulatory cells. Immunol Rev 2001;182:207–14. | Article | PubMed | ISI | CAS |
  4. Sakaguchi S, Sakaguchi N, Asano M, Itoh M, Toda M. Immunologic self-tolerance maintained by activated T cells expressing IL-2 receptor alpha-chains (CD25). Breakdown of a single mechanism of self-tolerance causes various autoimmune diseases. J Immunol 1995; 155: 1151–1164. | PubMed | ISI | CAS |
  5. Hori S, Nomura T, Sakaguchi S. Control of regulatory T cell development by the transcription factor Foxp3. Science 2003; 299: 1057–1061. | Article | PubMed | ISI | CAS |
  6. Liu Z, Tugulea S, Cortesini R, Suciu-Foca N. Specific suppression of T helper alloreactivity by allo-MHC class I-restricted CD8+CD28 T cells. Int Immunol 1998; 10: 775–783. | Article | PubMed | ISI | CAS |
  7. Jiang H, Ware R, Stall A, Flaherty L, Chess L, Pernis B. Murine CD8+ T cells that specifically delete autologous CD4+ T cells expressing V beta 8 TCR: a role of the Qa-1 molecule. Immunity 1995; 2: 185–194. | Article | PubMed | ISI | CAS |
  8. Abbas AK, Benoist C, Bluestone JA, Campbell DJ, Ghosh S, Hori S et al. Regulatory T cells: recommendations to simplify the nomenclature. Nat Immunol 2013; 14: 307–308. | Article | PubMed | ISI | CAS |
  9. Zheng SG, Gray JD, Ohtsuka K, Yamagiwa S, Horwitz DA. Generation ex vivo of TGF-beta-producing regulatory T cells from CD4+CD25 precursors. J Immunol 2002; 169: 4183–4189. | Article | PubMed | ISI | CAS |
  10. You S, Leforban B, Garcia C, Bach JF, Bluestone JA, Chatenoud L. Adaptive TGF-beta-dependent regulatory T cells control autoimmune diabetes and are a privileged target of anti-CD3 antibody treatment. Proc Natl Acad Sci USA 2007; 104: 6335–6340. | Article | PubMed | CAS |
  11. Rudensky AY. Regulatory T cells and Foxp3. Immunol Rev 2011; 241: 260–268. | Article | PubMed | ISI | CAS |
  12. Tone M, Tone Y, Adams E, Yates SF, Frewin MR, Cobbold SP et al. Mouse glucocorticoid-induced tumor necrosis factor receptor ligand is costimulatory for T cells. Proc Natl Acad Sci USA 2003; 100: 15059–15064. | Article | PubMed | CAS |
  13. Takahashi T, Tagami T, Yamazaki S, Uede T, Shimizu J, Sakaguchi N et al. Immunologic self-tolerance maintained by CD25+CD4+ regulatory T cells constitutively expressing cytotoxic T lymphocyte-associated antigen 4. J Exp Med 2000; 192: 303–310. | Article | PubMed | ISI | CAS |
  14. Miyara M, Yoshioka Y, Kitoh A, Shima T, Wing K, Niwa A et al. Functional delineation and differentiation dynamics of human CD4+ T cells expressing the FoxP3 transcription factor. Immunity 2009; 30: 899–911. | Article | PubMed | ISI | CAS |
  15. Bennett CL, Christie J, Ramsdell F, Brunkow ME, Ferguson PJ, Whitesell L et al. The immune dysregulation, polyendocrinopathy, enteropathy, X-linked syndrome (IPEX) is caused by mutations of FOXP3. Nat Genet 2001; 27: 20–21. | Article | PubMed | ISI | CAS |
  16. Brunkow ME, Jeffery EW, Hjerrild KA, Paeper B, Clark LB, Yasayko SA et al. Disruption of a new forkhead/winged-helix protein, scurfin, results in the fatal lymphoproliferative disorder of the scurfy mouse. Nat Genet 2001; 27: 68–73. | Article | PubMed | ISI | CAS |
  17. Tang Q, Bluestone JA. Regulatory T-cell therapy in transplantation: moving to the clinic. Cold Spring Harbor Perspect Med 2013; 3(11).
  18. Fontenot JD, Gavin MA, Rudensky AY. Foxp3 programs the development and function of CD4+CD25+ regulatory T cells. Nat Immunol 2003; 4: 330–336. | Article | PubMed | ISI | CAS |
  19. Williams LM, Rudensky AY. Maintenance of the Foxp3-dependent developmental program in mature regulatory T cells requires continued expression of Foxp3. Nat Immunol 2007; 8: 277–284. | Article | PubMed | ISI | CAS |
  20. Kim J, Lahl K, Hori S, Loddenkemper C, Chaudhry A, deRoos P et al. Cutting edge: depletion of Foxp3+ cells leads to induction of autoimmunity by specific ablation of regulatory T cells in genetically targeted mice. J Immunol 2009; 183: 7631–7634. | Article | PubMed | ISI |
  21. Fontenot JD, Rasmussen JP, Williams LM, Dooley JL, Farr AG, Rudensky AY. Regulatory T cell lineage specification by the forkhead transcription factor foxp3. Immunity 2005; 22: 329–341. | Article | PubMed | ISI | CAS |
  22. Koch MA, Tucker-Heard G, Perdue NR, Killebrew JR, Urdahl KB, Campbell DJ. The transcription factor T-bet controls regulatory T cell homeostasis and function during type 1 inflammation. Nat Immunol 2009; 10: 595–602. | Article | PubMed | ISI | CAS |
  23. Zheng Y, Chaudhry A, Kas A, deRoos P, Kim JM, Chu TT et al. Regulatory T-cell suppressor program co-opts transcription factor IRF4 to control TH2 responses. Nature 2009; 458: 351–356. | Article | PubMed | ISI | CAS |
  24. Chung Y, Tanaka S, Chu F, Nurieva RI, Martinez GJ, Rawal S et al. Follicular regulatory T cells expressing Foxp3 and Bcl-6 suppress germinal center reactions. Nat Med 2011; 17: 983–988. | Article | PubMed | ISI | CAS |
  25. Linterman MA, Pierson W, Lee SK, Kallies A, Kawamoto S, Rayner TF et al. Foxp3+ follicular regulatory T cells control the germinal center response. Nat Med 2011; 17: 975–982. | Article | PubMed | ISI | CAS |
  26. Chaudhry A, Rudra D, Treuting P, Samstein RM, Liang Y, Kas A et al. CD4+ regulatory T cells control TH17 responses in a Stat3-dependent manner. Science 2009; 326: 986–991. | Article | PubMed | ISI | CAS |
  27. Tontonoz P, Spiegelman BM. Fat and beyond: the diverse biology of PPARgamma. Annu Rev Biochem 2008; 77: 289–312. | Article | PubMed | ISI | CAS |
  28. Cipolletta D, Feuerer M, Li A, Kamei N, Lee J, Shoelson SE et al. PPAR-gamma is a major driver of the accumulation and phenotype of adipose tissue Treg cells. Nature 2012; 486: 549–553. | Article | PubMed | ISI | CAS |
  29. Burzyn D, Kuswanto W, Kolodin D, Shadrach JL, Cerletti M, Jang Y et al. A special population of regulatory T cells potentiates muscle repair. Cell 2013; 155: 1282–1295. | Article | PubMed | ISI | CAS |
  30. Shafiani S, Tucker-Heard G, Kariyone A, Takatsu K, Urdahl KB. Pathogen-specific regulatory T cells delay the arrival of effector T cells in the lung during early tuberculosis. J Exp Med 2010; 207: 1409–1420. | Article | PubMed | ISI | CAS |
  31. Shafiani S, Dinh C, Ertelt JM, Moguche AO, Siddiqui I, Smigiel KS et al. Pathogen-specific Treg cells expand early during mycobacterium tuberculosis infection but are later eliminated in response to interleukin-12. Immunity 2013; 38: 1261–1270. | Article | PubMed | ISI | CAS |
  32. Scott-Browne JP, Shafiani S, Tucker-Heard G, Ishida-Tsubota K, Fontenot JD, Rudensky AY et al. Expansion and function of Foxp3-expressing T regulatory cells during tuberculosis. J Exp Med 2007; 204: 2159–2169. | Article | PubMed | ISI | CAS |
  33. Facciabene A, Motz GT, Coukos G. T-regulatory cells: key players in tumor immune escape and angiogenesis. Cancer Res 2012; 72: 2162–2171. | Article | PubMed | ISI | CAS |
  34. Zhou X, Bailey-Bucktrout SL, Jeker LT, Penaranda C, Martinez-Llordella M, Ashby M et al. Instability of the transcription factor Foxp3 leads to the generation of pathogenic memory T cells in vivo. Nat Immunol 2009; 10: 1000–1007. | Article | PubMed | ISI | CAS |
  35. Yang XO, Nurieva R, Martinez GJ, Kang HS, Chung Y, Pappu BP et al. Molecular antagonism and plasticity of regulatory and inflammatory T cell programs. Immunity 2008; 29: 44–56. | Article | PubMed | ISI | CAS |
  36. Setoguchi R, Hori S, Takahashi T, Sakaguchi S. Homeostatic maintenance of natural Foxp3+ CD25+ CD4+ regulatory T cells by interleukin (IL)-2 and induction of autoimmune disease by IL-2 neutralization. J Exp Med 2005; 201: 723–735. | Article | PubMed | ISI | CAS |
  37. Miyao T, Floess S, Setoguchi R, Luche H, Fehling HJ, Waldmann H et al. Plasticity of Foxp3+ T cells reflects promiscuous Foxp3 expression in conventional T cells but not reprogramming of regulatory T cells. Immunity 2012; 36: 262–275. | Article | PubMed | ISI | CAS |
  38. Rudra D, Egawa T, Chong MM, Treuting P, Littman DR, Rudensky AY. Runx-CBFbeta complexes control expression of the transcription factor Foxp3 in regulatory T cells. Nat Immunol 2009; 10: 1170–1177. | Article | PubMed | ISI | CAS |
  39. Zheng Y, Josefowicz S, Chaudhry A, Peng XP, Forbush K, Rudensky AY. Role of conserved non-coding DNA elements in the Foxp3 gene in regulatory T-cell fate. Nature 2010; 463: 808–812. | Article | PubMed | ISI | CAS |
  40. Birzele F, Fauti T, Stahl H, Lenter MC, Simon E, Knebel D et al. Next-generation insights into regulatory T cells: expression profiling and FoxP3 occupancy in human. Nucleic Acids Res 2011; 39: 7946–7960. | Article | PubMed | ISI | CAS |
  41. Ono M, Yaguchi H, Ohkura N, Kitabayashi I, Nagamura Y, Nomura T et al. Foxp3 controls regulatory T-cell function by interacting with AML1/Runx1. Nature 2007; 446: 685–689. | Article | PubMed | ISI | CAS |
  42. Feng Y, Arvey A, Chinen T, van der Veeken J, Gasteiger G, Rudensky AY. Control of the inheritance of regulatory T cell identity by a cis element in the Foxp3 locus. Cell 2014; 158: 749–763. | Article | PubMed | ISI | CAS |
  43. Li X, Liang Y, LeBlanc M, Benner C, Zheng Y. Function of a Foxp3 cis-element in protecting regulatory T cell identity. Cell 2014; 158: 734–748. | Article | PubMed | ISI |
  44. Guo J, Zhang J, Zhang X, Zhang Z, Wei X, Zhou X. Constitutive activation of MEK1 promotes Treg cell instability in vivo. J Biol Chem 2014; 289: 35139–35148. | Article | PubMed | ISI |
  45. Bailey-Bucktrout SL, Martinez-Llordella M, Zhou X, Anthony B, Rosenthal W, Luche H et al. Self-antigen-driven activation induces instability of regulatory T cells during an inflammatory autoimmune response. Immunity 2013; 39: 949–962. | Article | PubMed | ISI | CAS |
  46. Shrestha S, Yang K, Guy C, Vogel P, Neale G, Chi H. Treg cells require the phosphatase PTEN to restrain TH1 and TFH cell responses. Nat Immunol 2015; 16: 178–187. | Article | PubMed |
  47. Huynh A, DuPage M, Priyadharshini B, Sage PT, Quiros J, Borges CM et al. Control of PI(3) kinase in Treg cells maintains homeostasis and lineage stability. Nat Immunol 2015; 16: 188–196. | Article | PubMed |
  48. Morawski PA, Mehra P, Chen C, Bhatti T, Wells AD. Foxp3 protein stability is regulated by cyclin-dependent kinase 2. J Biol Chem 2013; 288: 24494–24502. | Article | PubMed | ISI | CAS |
  49. Li Z, Lin F, Zhuo C, Deng G, Chen Z, Yin S et al. PIM1 kinase phosphorylates the human transcription factor FOXP3 at serine 422 to negatively regulate its activity under inflammation. J Biol Chem 2014; 289: 26872–26881. | Article | PubMed | ISI | CAS |
  50. Jordan MS, Boesteanu A, Reed AJ, Petrone AL, Holenbeck AE, Lerman MA et al. Thymic selection of CD4+CD25+ regulatory T cells induced by an agonist self-peptide. Nat Immunol 2001; 2: 301–306. | Article | PubMed | ISI | CAS |
  51. Pacholczyk R, Ignatowicz H, Kraj P, Ignatowicz L. Origin and T cell receptor diversity of Foxp3+CD4+CD25+ T cells. Immunity 2006; 25: 249–259. | Article | PubMed | ISI | CAS |
  52. Wong J, Obst R, Correia-Neves M, Losyev G, Mathis D, Benoist C. Adaptation of TCR repertoires to self-peptides in regulatory and nonregulatory CD4+ T cells. J Immunol 2007; 178: 7032–7041. | Article | PubMed | ISI | CAS |
  53. Hsieh CS, Liang Y, Tyznik AJ, Self SG, Liggitt D, Rudensky AY. Recognition of the peripheral self by naturally arising CD25+ CD4+ T cell receptors. Immunity 2004; 21: 267–277. | Article | PubMed | ISI | CAS |
  54. Facciabene A, Peng X, Hagemann IS, Balint K, Barchetti A, Wang LP et al. Tumour hypoxia promotes tolerance and angiogenesis via CCL28 and Treg cells. Nature 2011; 475: 226–230. | Article | PubMed | ISI | CAS |
  55. Tan W, Zhang W, Strasner A, Grivennikov S, Cheng JQ, Hoffman RM et al. Tumour-infiltrating regulatory T cells stimulate mammary cancer metastasis through RANKL–RANK signalling. Nature 2011; 470: 548–553. | Article | PubMed | ISI | CAS |
  56. Blatner NR, Mulcahy MF, Dennis KL, Scholtens D, Bentrem DJ, Phillips JD et al. Expression of RORgammat marks a pathogenic regulatory T cell subset in human colon cancer. Sci Transl Med 2012; 4: 164ra59. | Article |
  57. Rivers TM. Viruses and Koch’s Postulates. J Bacteriol 1937; 33: 1–12. | PubMed | CAS |
  58. Xu L, Kitani A, Fuss I, Strober W. Cutting edge: regulatory T cells induce CD4+CD25Foxp3 T cells or are self-induced to become Th17 cells in the absence of exogenous TGF-beta. J Immunol 2007; 178: 6725–6729. | Article | PubMed | ISI | CAS |
  59. Zheng SG, Wang J, Horwitz DA. Cutting edge: Foxp3+CD4+CD25+ regulatory T cells induced by IL-2 and TGF-beta are resistant to Th17 conversion by IL-6. J Immunol 2008; 180: 7112–7116. | Article | PubMed | ISI | CAS |
  60. Duarte JH, Zelenay S, Bergman ML, Martins AC, Demengeot J. Natural Treg cells spontaneously differentiate into pathogenic helper cells in lymphopenic conditions. Eur J Immunol 2009; 39: 948–955. | Article | PubMed | ISI | CAS |
  61. Zhou X, Jeker LT, Fife BT, Zhu S, Anderson MS, McManus MT et al. Selective miRNA disruption in T reg cells leads to uncontrolled autoimmunity. J Exp Med 2008; 205: 1983–1991. | Article | PubMed | ISI | CAS |
  62. Srinivas S, Watanabe T, Lin CS, William CM, Tanabe Y, Jessell TM et al. Cre reporter strains produced by targeted insertion of EYFP and ECFP into the ROSA26 locus. BMC Dev Biol 2001; 1: 4. | Article | PubMed | CAS |
  63. Komatsu N, Okamoto K, Sawa S, Nakashima T, Oh-hora M, Kodama T et al. Pathogenic conversion of Foxp3+ T cells into TH17 cells in autoimmune arthritis. Nat Med 2014; 20: 62–68. | Article | PubMed | ISI | CAS |
  64. Sato K, Suematsu A, Okamoto K, Yamaguchi A, Morishita Y, Kadono Y et al. Th17 functions as an osteoclastogenic helper T cell subset that links T cell activation and bone destruction. J Exp Med 2006; 203: 2673–2682. | Article | PubMed | ISI | CAS |
  65. Deknuydt F, Bioley G, Valmori D, Ayyoub M. IL-1beta and IL-2 convert human Treg into TH17 cells. Clin Immunol 2009; 131: 298–307. | Article | PubMed | ISI | CAS |
  66. Koenen HJ, Smeets RL, Vink PM, van Rijssen E, Boots AM, Joosten I. Human CD25highFoxp3pos regulatory T cells differentiate into IL-17-producing cells. Blood 2008; 112: 2340–2352. | Article | PubMed | ISI | CAS |
  67. Bovenschen HJ, van de Kerkhof PC, van Erp PE, Woestenenk R, Joosten I, Koenen HJ. Foxp3+ regulatory T cells of psoriasis patients easily differentiate into IL-17A-producing cells and are found in lesional skin. J Invest Dermatol 2011; 131: 1853–1860. | Article | PubMed | ISI |
  68. Takahashi R, Nishimoto S, Muto G, Sekiya T, Tamiya T, Kimura A et al. SOCS1 is essential for regulatory T cell functions by preventing loss of Foxp3 expression as well as IFN-gamma and IL-17A production. J Exp Med 2011; 208: 2055–2067. | Article | PubMed | ISI | CAS |
  69. Takimoto T, Wakabayashi Y, Sekiya T, Inoue N, Morita R, Ichiyama K et al. Smad2 and Smad3 are redundantly essential for the TGF-beta-mediated regulation of regulatory T plasticity and Th1 development. J Immunol 2010; 185: 842–855. | Article | PubMed | ISI | CAS |
  70. Ouyang W, Liao W, Luo CT, Yin N, Huse M, Kim MV et al. Novel Foxo1-dependent transcriptional programs control Treg cell function. Nature 2012; 491: 554–559. | Article | PubMed | ISI | CAS |
  71. Muto G, Kotani H, Kondo T, Morita R, Tsuruta S, Kobayashi T et al. TRAF6 is essential for maintenance of regulatory T cells that suppress Th2 type autoimmunity. PloS One 2013; 8: e74639. | Article | PubMed |
  72. Wan YY, Flavell RA. Regulatory T-cell functions are subverted and converted owing to attenuated Foxp3 expression. Nature 2007; 445: 766–770. | Article | PubMed | ISI | CAS |
  73. Onizuka S, Tawara I, Shimizu J, Sakaguchi S, Fujita T, Nakayama E. Tumor rejection by in vivo administration of anti-CD25 (interleukin-2 receptor alpha) monoclonal antibody. Cancer Res 1999; 59: 3128–3133. | PubMed | ISI | CAS |
  74. Shimizu J, Yamazaki S, Takahashi T, Ishida Y, Sakaguchi S. Stimulation of CD25+CD4+ regulatory T cells through GITR breaks immunological self-tolerance. Nat Immunol 2002; 3: 135–1342. | Article | PubMed | ISI | CAS |
  75. Peggs KS, Quezada SA, Chambers CA, Korman AJ, Allison JP. Blockade of CTLA-4 on both effector and regulatory T cell compartments contributes to the antitumor activity of anti-CTLA-4 antibodies. J Exp Med 2009; 206: 1717–1725. | Article | PubMed | ISI | CAS |
  76. Piconese S, Valzasina B, Colombo MP. OX40 triggering blocks suppression by regulatory T cells and facilitates tumor rejection. J Exp Med 2008; 205: 825–839. | Article | PubMed | ISI | CAS |
  77. Yamaguchi T, Hirota K, Nagahama K, Ohkawa K, Takahashi T, Nomura T et al. Control of immune responses by antigen-specific regulatory T cells expressing the folate receptor. Immunity 2007; 27: 145–159. | Article | PubMed | ISI | CAS |
  78. Liu Y, Wang L, Predina J, Han R, Beier UH, Wang LC et al. Inhibition of p300 impairs Foxp3+ T regulatory cell function and promotes antitumor immunity. Nat Med 2013; 19: 1173–1177. | Article | PubMed | ISI | CAS |
  79. Lu L, Lan Q, Li Z, Zhou X, Gu J, Li Q et al. Critical role of all-trans retinoic acid in stabilizing human natural regulatory T cells under inflammatory conditions. Proc Natl Acad Sci USA 2014; 111: E3432–E3440. | Article | PubMed |
  80. Zhou X, Kong N, Wang J, Fan H, Zou H, Horwitz D et al. Cutting edge: all-trans retinoic acid sustains the stability and function of natural regulatory T cells in an inflammatory milieu. J Immunol 2010; 185: 2675–2679. | Article | PubMed | ISI | CAS |


This work was supported by the National Key Basic Research and Development 973 Program of China (Grant No. 2012CB917102), the National Natural Science Foundation of China (Grant No. 31270959) and a grant from the China Postdoctoral Science Foundation (No 2011M500422).