Molecular mechanism for impaired suppressive function of Tregs in autoimmune diseases: A summary of cell‐intrinsic and cell‐extrinsic factors

Abstract Regulatory T (Treg) cells are responsible for maintaining immune homeostasis and preventing autoimmunity. In immune homeostasis condition, Tregs exert their suppressive function through inhibiting the proliferation of effector T cells. In response to environmental signals, Tregs display phenotypic heterogeneity and altered stability, which endows their suppressive function in a context‐dependent manner. Compelling evidence indicates deficiency of Treg suppressive function is related to the immunopathogenesis of various autoimmune diseases. Consequently, it is vital to further our understanding of the molecular mechanism accounting for the regulation of Treg suppressive functions. In this review, we outline the current knowledge that highlights how cell‐intrinsic factors, such as inflammatory cytokines, transcription factors, signalling pathways, post‐translational modification (PTM), miRNAs, protein and protein complex, and cell‐extrinsic factors orchestrate the suppressive function of Tregs. Improved understanding of the molecular mechanism related to the suppressive functional property of Tregs should provide new insights into autoimmunity and disease pathogenesis, which offers opportunity for identifying new therapeutic targets for Treg‐related autoimmune diseases and cancers.

Epigenetic modification and post-translational modification of Foxp3 which reduced the expression of Foxp3 will impair the differentiation of Tregs and their suppressive function. 3 Thus, Foxp3 is regarded as the master transcription factor of Tregs.
Dysregulation of Tregs has been related to the pathogenesis of various autoimmune disease, such as psoriasis, rheumatoid arthritis (RA), systemic lupus erythematosus (SLE) and diabetes. 4 The impaired suppressive function of Tregs will lead to uncontrolled proliferation of effector T cells, which exaggerate the inflammatory process and contribute to disease pathogenesis. Therefore, it is of importance to illustrate the molecular mechanism accounting for dysfunctional Tregs. The present review summarizes various factors, including inflammatory cytokines, transcription factors, miRNAs, post-translational modifications and extrinsic factors such as impact by other cells. We do not include metabolic factors and regulation of Foxp3 by epigenetic modification as there are already detailed reviews in this field. Our purpose is to briefly summarize the current knowledge on the factors that influence the suppressive functions of Tregs and to provide theoretical basis for directing studies in the field of Tregs in different disease.

| Inflammatory cytokines
The microenvironment in patients with autoimmune diseases is very complex and encompasses a series of pro-inflammatory cytokines such as IL-6, TNF-α, IL-21 and IFN-γ. Despite the fact that these pro-inflammatory cytokines are critical in driving the plasticity of Tregs, they are capable of regulating the suppressive function of Tregs.
In vitro, IL-6 addition induced a significant decrease in both the proportion of CD4 + CD25 + Foxp3 + Tregs and their suppressive functions. 5 Moreover, in psoriasis, it is discovered that elevated IL-6 from endothelial cells, dendritic cells and Th17 cells induced phosphorylation of STAT3 in both Tregs and Teff cells, which dampened the suppressive function of Tregs to Teff cells. 6 IL-21, a CD4 + T cell-derived cytokine, is shown to enhance inflammatory response in autoimmune disease. In one study, it is suggested that IL-21 rendered CD4 + CD25 -T cells resistant to Tregmediated suppression, which impaired the suppressive function of Tregs. 7 Other mechanisms relating to IL-21-mediated suppression of Foxp3 have been proposed. One possibility might be that IL-21 could reduce the expression and stability of Foxp3 in CD4 + T cells. 8 By using IL-21R -/mice in asthma and colitis, Tortola et al 9 uncovered a direct effect of IL-21 on promoting apoptosis of Tregs. By further investigation, IL-21 was shown to interfere with expression of Bcl-2 family genes which sensitized these Tregs to apoptosis.
TNF-α has also been shown to regulate the inhibitory function of Tregs. However, contradictory effects of TNF-α on the function of Tregs have been reported. TNF-α is long considered to be potent pro-inflammatory cytokines that are implicated in the pathogenesis of various diseases; however, the pleiotropic effect of TNF-α such as anti-inflammatory function has been proposed. In rheumatoid arthritis (RA), TNF-α decreased the expression of Foxp3, which down-regulated the suppressive function of Tregs. Anti-TNF therapy with infliximab restored the suppressive function of Tregs through a mechanism involving an increase in Foxp3. 10 13 In graft-versus-host disease and tumour environment, TNF-α was shown to enhance the suppressive function of Tregs mainly through a TNF-α-TNFR2 signalling. 14 Other pro-inflammatory cytokines such as IL-1β was also capable of abrogating the suppressive function of Tregs. 15 Clarifying the complex crosstalk of these cytokines is critical to understand the mechanisms accounting for dysfunctional Tregs.  However, Tregs from EOS-deficient mice endow the same suppressive function as the control mice in vitro. 18 Another study using siRNA knockdown technology reported controversial results. Eos knockdown abrogated the suppressive function of Tregs both in vivo and in vitro. 19 Other positive regulator of Treg suppressive function, such as YAP, 20 has also been discovered. In summary, these transcription factors, either positively or negatively, function mainly through the following three mechanisms to regulate the function of Tregs. First, by binding to the Foxp3 locus, they are capable of regulating its expression. Secondly, they can act as partners of Foxp3 and physically bind to Foxp3 by protein-protein interaction, which may regulate its structure and induce chromatin modifications. Thirdly, some of these transcription factors such as Hhex bind to the promoters of Treg signature genes such as CTLA4 and IL2RA ( Figure 1).

| Signalling pathways
Among the identified signalling pathways, NF-κB shows its importance in maintaining Treg stability and suppressive function. NF-κB signalling is classified into two pathways: the canonical pathway which encompasses p50, c-Rel and p65 (RelA), and the non-canonical pathway which involves TNF receptor family members (TNFR), NF-κB-inducing kinase (NIK), p100 and RelB.
The canonical signalling is proven to be the mast regulator of Treg development and for the maintenance of Treg suppressive function. In mice, selective deletion of Rela in Tregs lost their ability to suppress the proliferation of Teff cells. Moreover, the secretion of inflammatory cytokines IL-17 and IFN-γ was increased. Deletion of both Rela and c-rel in Tregs failed to prevent the colitis of mouse models, which indicated that deletion of c-rel and Rela led to complete loss of Treg suppressive function in vivo and in vitro. 21 As constitutive TCR signalling is required for the maintenance of Treg suppressive function, it is reasonable that TCR signalling involved canonical activation of NF-κB pathway is important for Treg function.
The alternative non-canonical pathway of NF-κB shows great potential in regulating the Treg function and homeostasis. Conditional which is required for the non-canonical activation of NF-κB, is also required for the proliferative and suppressive function of Tregs.
IKKα-deficient Tregs are impaired in their proliferative and suppressive functions. In vivo, IKKα-deficient Tregs failed to prevent the development of colitis. 24 However, constitutive activation of NIK (NF-κB-inducing kinase), which links the TNFR to non-canonical activation of NF-κB, impaired the suppressive function of Tregs and endows Tregs with a pro-inflammatory phenotype. 25 Notch signalling, which is linked to canonical and non-canonical activation of NF-κB, is controversial in its role to regulate Treg activation and function. A few studies indicated the negative effect of Notch on Treg in RA, diabetes. In two arthritis mouse models, blockade of Notch1 increased the Treg population and suppressive ability. 26 Other studies suggested Notch signalling as positive regulators of Treg function. In these studies, Notch activation recruits complex to bind to Foxp3 promoter and activate its expression. 27 Thus, Notch controls Treg function and homeostasis in a context-dependent manner.
Other signalling, such as PI3K and TLR signalling, is also shown to be important for the maintenance of Treg suppressive function. 28,29 However, despite the fact that all the aforementioned signalling pathways are necessary for maintaining Treg suppressive function, the subtle difference in the level of activation such as hyperactivation of the signalling pathways can antagonize the suppressive function of Tregs.

| Post-translational modification (PTM)
Foxp3 is the master regulator of Treg development of suppressive activation. Previous reviews have already summarized that Foxp3 could be modulated by ubiquitination, phosphorylation, O-GlcNAcylation, acetylation, ubiquitylation and methylation. 30 Here we focused on the recent findings concerning the role of ubiquitination in the regulation of Treg function.
Although the best-known function of ubiquitination is to direct proteins for degradation, it regulates diverse biological functions which depends both on specific targeted residue on substrates and the specific linkages types that formed between ubiquitin-to-ubiquitin. One mechanism by which ubiquitination modulates the sup- miR-181a/b-deficient mice displayed elevated suppressive capacity in vivo. 38 Other mechanism, such as miR-568 induced inhibition of NFAT5, a transcription factor, is also involved in inhibiting the suppressive function of Tregs. 39 41 Molecular mechanism analysis showed four potential F I G U R E 3 Protein and protein complex mediated regulation of Treg immunosuppression. PAR4 involved activation of PI3K-AKT pathway to negatively regulate the transcription of Foxo1, and an inactivation of STAT5 pathway to co-ordinate a negative regulation on Foxp3 expression. Scaffold protein Bcl10 promotes the activation of NF-κB, the expression of Treg effector molecules such as KLRG1 and Icos, the master transcription factor Foxp3 and Treg-associated suppressive molecules like TGF-β1 and IL-10 to maintain the suppressive function of Tregs pathways that Bcl10 might involve to maintain the suppressive function of Tregs: Knockdown of Bcl10 decreased the activation of NF-κB, the expression of Treg effector molecules such as KLRG1 and Icos, the master transcription factor Foxp3 and Treg-associated suppressive molecules like TGF-β1 and IL-10 ( Figure 3).

| Protein and protein complex
Tregs exert immunosuppressive functions and prevent the development of autoimmune diseases. However, in tumour microenvironment, Tregs prevent anti-cancer immunity by suppressing antitumour effector cells. Thus, in this specific tumour context, decreasing the suppressive function will favour antitumour therapies. Neuropilin-1 (NRP-1), a non-tyrosine kinase receptor, is preferentially expressed in intratumoural Tregs. NRP1 interacts with semaphorin 4a and forms a complex with VEGFR2 to potentiate the Treg function and survival.
Antagonist targeting NRP1 significantly reduced Foxp3 stability and enhanced IFN-γ production, which dampened the suppressive function of intratumoural Tregs. 42   The earliest approach is by adoptive transfer of Tregs. However, the side effects and the Treg instability limit the widespread application of this therapy. 49 More recent trials for Treg immunotherapy have switched to IL-2-induced expansion of Tregs. As IL-2 receptor is not specifically expressed in Tregs, IL-2 stimulation can activate a range of IL-2 responsive cells including CD4 + and CD8 + effector T cells and NK cells. Thus, the challenge is how to reduce the binding of IL-2 to other non-Treg cells. As Tregs expressed all three components of the IL-2 receptor: αβγ, which differed from other cells expressing only IL-2Rβγ, one optimal solution is to reduce IL-2 affinity to beta-gamma chain. For this, Peterson et al 50 developed a long-lived bivalent fusion protein IgG-(IL-2N88D)2, which showed reduced affinity for IL-2Rβγ. Treatment of cynomolgus monkeys with low doses of this IL-2 mutein protein allowed for a sustained 10-to 14-fold increase in CD4 + and CD8 + Tregs, which opens up possibility for using IL-2 therapy to treat autoimmune diseases. Multiple approaches that expand ex vivo Tregs in autoimmune disease are under investigation, but the carriers that could deliver these Tregs in vivo are the challenge. Nanoparticles (NPs) encapsulating peptides represent a potential modality to carry Tregs. It is reported that NPs encapsulating IL-2 and TGF-β coated with anti-CD2/CD4 antibodies resulted in an expansion of Tregs in vivo and alleviated SLE phenotypes in mouse models. 51 Although multiple approaches have been developed to expand Treg population, it is of vital importance to take consideration of Treg stability, Treg specificity and side effects into account to obtain an optimal approach for the treatment of autoimmune disease. In tumour microenvironment (TME), Tregs are a negative regulator as it suppresses antitumour immunity. In cancer, Treg-targeted therapy is either to down-regulate the suppressive function of Tregs or to deplete Treg population. Tregs in the TME express the IC molecules like CTLA-4 and PD-1, chemokines receptor like CCR4 and CCR8, and some proteins like Nrp1, can be utilized to deplete Tregs in TME. For example, anti-PD-1 therapy has been largely investigated in the treatment of cancer patients. Anti-PD-1 antibody is efficient in decreasing intratumoural Tregs and suppress tumour volume and tumour growth. 52 Compared with traditional anti-CTLA4 therapy, recently developed tumour-conditional anti-CTLA4 therapy depleted tumour-infiltrated Tregs, while preserving tissue-resident Tregs, which preserved antitumour effects and reduced multiorgan immune toxicity. 53 To sum up, developing novel approaches to specifically target Tregs will represent next-generation therapy for the treatment of autoimmune disease and cancers.

ACK N OWLED G EM ENT
This work was supported by grants from the National Natural Science Foundation of China (grant number 81773265).

CO N FLI C T O F I NTE R E S T S
The authors declare no competing interests.