Regulatory T cells in acute and chronic kidney diseases

Abstract

Foxp3-expressing CD4⁺ regulatory T cells (Tregs) make up one subset of the helper T cells (Th) and are one of the major mechanisms of peripheral tolerance. Tregs prevent abnormal activation of the immune system throughout the lifespan, thus protecting from autoimmune and inflammatory diseases. Recent studies have elucidated the role of Tregs beyond autoimmunity. Tregs play important functions in controlling not only innate and adaptive immune cell activation, but also regulate nonimmune cell function during insults and injury. Inflammation contributes to a multitude of acute and chronic diseases affecting the kidneys. This review examines the role of Tregs in pathogenesis of renal inflammatory diseases and explores the approaches for enhancing Tregs for prevention and therapy of renal inflammation.

INTRODUCTION

The immune system is a double-edged sword, which on one hand performs immunosurveillance to protect from pathogenic invaders such as viruses, bacteria, fungi, etc., as well as internal threats such as cancers and neoplasm, but on the other hand is involved in clearance and recycling of dying cells. The immune system is armed with checkpoints, which ensure that once the pathogen has been eliminated, the immune system returns to a quiescent basal state. The immune system exerts these functions while distinguishing what is self and what is not. The “self” includes the harmless commensal organisms, which do not activate the immune system and are “tolerated.”

The two components of the immune system, namely innate and adaptive immunity, use diverse mechanisms to identify threats. The innate immune cells such as the phagocytes (primarily granulocytes, macrophages, and dendritic cells) identify conserved pathogen associated molecular patterns or danger associated molecular patterns (DAMPs) to identify the threat (144, 157). The adaptive immune cells on the other hand are educated in primary lymphoid organs to differentiate self and foreign antigens through a mechanism categorized as central tolerance, where any developing T and B cells die through apoptosis if their T-cell receptors (TCRs) or B-cell receptors (BCRs) react strongly to self-antigens (185, 245). Only B cells whose BCR does not react to self-antigen or T cells whose TCR reacts moderately to the self-antigen presented on the major histocompatibility complex (MHC) or human leukocyte antigen survive and are exported to the periphery. T cells responding to the antigen on MHC I give rise to CD8⁺ cytotoxic T cells, whereas T cells antigen recognition on MHC II generates CD4⁺ helper T cells (Th).

RECOGNITION AND IDENTITY OF REGULATORY T CELLS

A subset of Th cells, which react strongly to self-antigen-MHC II complex, does not get eliminated and is rather positively selected to be exported to periphery and regulate the activation of other immune cells. These cells express the forkhead box p3 (Foxp3) transcription factor and have been christened regulatory T cells (Tregs) (92, 170, 217). Cells with similar regulatory properties were first recognized in mice decades ago, when it was found that removal of thymus between days 2 and 4 after birth results in autoimmune diseases, and replenishing the thymocytes in these day-3 thymectomized mice prevented the development of autoimmunity, suggesting that certain T cells emigrate from the thymus of postnatal animals, which have the ability to “suppress” autoimmunity (160, 166).

Later work from the laboratories of Sakaguchi, Shevach, and others identified the suppressor cells in the thymus as well as periphery to express high levels of interleukin (IL)-2 high-affinity receptor (IL-2Rα or CD25) as CD4⁺CD25^hi cells (196, 237). Subsequently, in the early 2000s Foxp3 was identified as a lineage-defining factor and a marker to confidently identify Tregs, based on studies on mutations in the Foxp3 gene in mice (scurfy) and humans [immune dysregulation, polyendocrinopathy, enteropathy, and X-linked (IPEX) syndrome] (14, 254). Absence of functional Foxp3 induces a multiorgan inflammatory syndrome and death in infancy. Early replenishment with CD4⁺CD25⁺ cells, the majority of which also express Foxp3, from normal mice prevented the mortality and autoimmune syndrome, indicating that the CD4⁺CD25⁺ Tregs possess in their repertoire, the ability to suppress multiorgan inflammation (21, 210, 223). Fontenot et al. (57) and Williams and Rudensky (255) generated mice with targeted deletion of Foxp3 and Hori et al. (77) made Foxp3 overexpressing mice and confirmed these findings. Owing to their specificity to self-antigen and constant exposure to self-antigen, Foxp3⁺ Tregs express the properties of activated antigen-experienced cells including high expression of CD44 and CD25 (IL-2Rα) (54, 59, 181, 195).

CD25 is not merely a surface marker for Tregs, but the survival and function of Tregs is also critically dependent on IL-2 (4, 63, 237, 242). Similar to the deficiency of Foxp3, lack of IL-2/IL-2R causes multiorgan inflammatory disease and death in infancy (4, 153, 259). CD25 expression is also upregulated on activated non-Treg cells, although not to the same extent as on Tregs, however, making it harder to distinguish Tregs from activated T cells. Fluorescent reporters for Foxp3 expression have been generated in mice, thus enabling sophisticated studies (58, 74). Such approaches are not possible in human studies. However, inclusion of IL-7Rα (CD127) has helped distinguish Tregs from activated T cells, such that CD4⁺CD25^hiCD127^lo cells are widely accepted as Tregs with more than 95% of these cells expressing Foxp3 (205).

Attempts have also been made to distinguish the thymic-derived Tregs (tTregs) from peripherally-induced Tregs (pTregs). Thornton et al. (236) postulated the expression of Helios transcription factor to differentiates tTregs from pTregs, such that the proportion of Helios⁺ Tregs is higher in thymus than periphery, with the proportion of Helios⁺ Tregs declining in the periphery with age. Helios was also found to regulate the fitness of CD44⁺CD62L^lo effector Tregs. Although there was no overt pathology of Treg-specific deletion of Helios, such Tregs had impaired ability to regulate activation of T cells and germinal center (GC) responses (204). Other cell surface markers have been documented to differentiate the tTregs and pTregs including the “T cell immunoreceptor with Ig and ITIM domain” (TIGIT), FcR-like 3 (FCRL3), Neuropilin-1 (Nrp1), etc. (17, 265, 272), with some controversy (221, 228).

SUBSETS OF TREGs

The Foxp3 Tregs generated during T-cell selection in thymus are commonly known as thymus-derived Tregs (tTregs) or natural Tregs (nTregs). Tregs not only regulate immune response to self-antigen but also play an important role in maintaining tolerance to commensal organisms, food, and air-borne antigens as well as the fetus, which essentially is a semi-allograft (6, 80, 91). Foxp3⁺ Tregs are also generated from naïve T cells during antigenic response to nonself or neoantigens in the presence of transforming growth factor (TGF)-β and IL-2 and are called induced Tregs (iTregs), peripherally-derived Tregs (pTregs), or adaptive Tregs (aTregs) in the literature (31, 116, 122). The majority of iTregs reside at the surfaces that are frequently exposed to the environment, such as skin, mucosa, or placenta. Other subsets of Tregs have been defined that do not express Foxp3, yet are immunosuppressive and produce IL-10, TGF-β, or IL-35 and are termed Tr1, Th3, or Tr35, respectively (25, 36, 37, 71, 169). Regulatory cells other than those belonging to Th lineage have also been described. These include CD4⁻CD8⁻ double-negative Tregs, Qa-1-restricted CD8 Tregs, CD28⁺CD8⁺ Tregs, regulatory NK cells, regulatory B cells, etc. (150, 191, 218, 234, 247, 249). For the purpose of this review, we will focus our discussion on the CD4⁺Foxp3⁺ Tregs only.

ESSENTIAL REQUIREMENTS FOR TREG GENERATION AND FUNCTION

Although several pathways have been identified for the generation, survival, and function of Tregs (Fig. 1), TCR and IL-2 signaling remain as the minimum requirements for generation of Tregs in the thymus. We discuss these and other mechanisms below.

Fig. 1.

Regulators and regulatory mechanisms of Tregs. The CD4⁺Foxp3⁺ Tregs are primarily defined by the expression of Foxp3 transcription factor and high-affinity IL-2 receptor (IL-2Rα or CD25). The most important factors for thymic Tregs, which also express Helios transcription factor, include TCR/CD28 signaling and IL-2. In addition, several other mechanisms are utilized for the differentiation, maintenance, and functions of Tregs. TGF-β, which is one of the suppression mechanisms used by Tregs, along with IL-2, converts naïve T cells to Tregs during T-cell receptor (TCR) stimulation. IL-33 is the newest factor that can promote differentiation, proliferation, and function of Tregs. Other major mechanisms are highlighted. Tregs can suppress target cells through IL-2 consumption, adenosine, soluble mediators, costimulatory molecules, and direct cytotoxicity. Trafficking and adhesion molecules also play important roles in Treg function. Costimulatory molecules: CD28; CTLA-4, cytotoxic T-lymphocyte-associated protein 4; TIGIT, T-cell immunoreceptor with Ig and ITIM domains; GARP’, glycoprotein A repetitions predominant; GITR, glucocorticoid-induced tumor necrosis factor-related receptor; PD-1, programmed death 1. Cytokines: IL-2, interleukin-2; tripartite IL-2 receptor IL-2Rα, CD25, IL-2Rβ, and IL-2Rγ; ST2, IL-33 receptor; interleukin (IL)-2 receptor α chain IL-33, interleukin-33; 2A_2AR; TGF-β, transforming growth factor-β; TGF-βR, TGFβ receptor. Soluble mediators of suppression: IL-10, interleukin-10; IL-35, interleukin-35. Cytotoxicity: granzyme B; FasL, Fas ligand. Adhesion molecules: ItgαVβ8, integrin αVβ8 (also involved in TGF-β activation); CD103, integrin αE; Nrp1, neuropilin-1. Adenosine metabolism: CD39 and CD73, ectonucleotidases; A2A adenosine receptor.

T-Cell Receptor

The first identification that antigen specificity is essential for generation of Tregs came from TCR-transgenic mice (6, 91, 248), with specificity to a non-self-antigen (e.g., ovalbumin). In these mice, the majority of T cells express a rearranged TCR for ovalbumin (OVA). However, Foxp3-expressing Tregs in such mice were only found in the minor population of cells not bearing the transgenic TCR. On a recombination (Rag1)- deficient background the endogenous TCR was not expressed, resulting in a complete lack of Tregs. When the cognate antigen (ovalbumin) was introduced as a second transgene, large number of CD4⁺ T cells were deleted, but a sizable proportion of the remaining OVA-specific T cells became Tregs. Interestingly, in an OTII-TCR-transgenic Foxp3 mutant mice (scurfy), it was primarily the nontransgenic TCR-bearing cells that expanded in the target organs and contributed to the pathology (211). Similarly, T cells engineered to express tTreg-derived TCRs expanded more in the lymphopenic hosts, indicating their greater self-reactivity (123).

Persistent TCR signaling is not only important for generation of Tregs in thymus, but sustained expression of TCR is necessary for maintaining their suppressive activity (203). By using a sophisticated double-transgenic mouse strain bearing Foxp3^{eGFP-Cre-ERT2} mated to loxp-flanked TCRα chain, Trac^FL/WT, or Trac^FL/FL mice, TCR expression can be deleted in the peripheral Tregs by tamoxifen treatment, and it was shown that although expression of Foxp3 was not affected, a specific subset of CD44^hiCD62L^lo effector-like Tregs was lost (131). This also correlated with developing fulminant autoimmunity in less than 2 wk of TCR deletion, indicating that sustained expression of TCR is critical for maintenance of Treg suppressive activity. Interestingly, in these TCR-deleted Tregs that expressed normal levels of CD25 and other Tregs associated markers (CD39, CD73, GITR), the responsiveness to IL-2 was maintained, thus suggesting that CD25 expression on Tregs is independent of continued TCR expression. However, it must be noted that Tregs once activated by one antigen can suppress proliferation of not only T-cell activation to a different antigen, but also in a xenogeneic manner (184).

IL-2/IL-2R

IL-2 was discovered more than three decades ago and was labeled as T-cell growth factor. However, studies with mice deficient in IL-2 or IL-2Rα (CD25) have shown that IL-2 is critical for the survival and maintenance of Tregs (33, 40, 58, 79) and may be dispensable for Th1, but not Th2 cells (212, 214). Mice deficient in IL-2 or IL-2R have a reduced proportion of Tregs, especially in the periphery (4, 63). As a result, the IL-2/IL-2R-deficient mice develop a lympho-proliferative disease and die young (208, 274). Interestingly, absence of CD25 expression results in Treg deficiency, which leads to polyclonal activation of T-cells and increased production of IL-2. The reduced consumption (likely by Tregs) and increased production of IL-2 leads to accumulation of very high levels of IL-2 (serum levels exceeding 1 mg/ml), which drives an expansion of memory phenotype CD8⁺ T cells (216). The mortality, IL-2 levels, and CD8 T-cell proliferation could be normalized by adoptive transfer of CD4⁺CD25⁺ Tregs from wild-type mice. STAT5 regulates Foxp3 expression in mice and humans (164, 260). Ablation of STAT5, which is downstream to IL-2 signaling also results in loss of Tregs in thymus as well as the periphery, with phenotypes similar to the IL-2/IL-2R-deficient mice (22, 268). Activated Stat5 binds to two different regulatory elements in the Foxp3 gene (see below) and has an important role in its expression, in addition to being a paracrine proliferative signal for Tregs. During thymopoiesis, CD4⁺CD25⁺Foxp3⁻ precursors of Tregs could be identified, which begin to express Foxp3 in response to IL-2 in vitro and in vivo (141, 215, 268). IL-2 not only regulates Tregs directly as a proliferation/maintenance factor and by increasing the expression of Foxp3 directly, but also regulates the responsiveness of Tregs to other stimuli through CD103 (213) and ST2 (see below).

Costimulation

Whereas the interaction of TCR with MHC-antigen complex provides “signal 1,” adequate T-cell activation depends on “signal 2” provided by the interaction of costimulatory molecules (B7-1 and 2, also known as CD80 and CD86) on the antigen-presenting cells (APCs) with their counterpart on the T cells (CD28). These signals are also important for Treg differentiation, since Treg selection is based on the strength of the activation signal. The proportion of Tregs was strongly reduced in mice with genetic deficiency of CD28 or CD80/CD86 (197, 229, 232, 248). Although CD28 signaling in conventional T cells is critical for IL-2 production, interaction of the COOH-terminal cytoplasmic tail of CD28 with “lck” plays an intrinsic role in inducing the expression of Foxp3, glucocorticoid-induced TNFR-related (GITR), and CTLA-4 on antigen-experienced double-positive thymocytes (229). Several transcription factors downstream to the TCR/CD28 signaling bind to the promoter as well as intronic regulatory regions of the Foxp3 gene, as further discussed below. CTLA-4 is another costimulatory factor expressed late on T cells after multiple rounds of stimulation that, similar to CD28, interacts with CD80/CD86 on the APCs (24, 117). Unlike the activation signal relayed by CD28, CTLA-4 is believed to send inhibitory signal in the T cell. Since Tregs are always activated in response to the ubiquitous self-antigen (186, 231), they also express CTLA-4, and in Tregs, CTLA-4 ligation to CD80/CD86 sends a signal to downmodulate the activation of APCs (187, 233). Global or Treg-specific ablation of CTLA-4 resulted in greater number of Tregs; however, it impaired the suppressive activity of Tregs and induced activation and proliferation of DC and inflammatory conditions (61, 178, 256). CTLA-4 mutations have been reported to have a high degree of correlation with several autoimmune diseases, including Type 1 diabetes and multiple sclerosis (30, 53, 199).

Programmed death 1 (PD-1), expressed on T cells and PD-L1/PD-L2 (B7-H1 and B7-H2), expressed on the APC as well as nonimmune cells is another pair of costimulatory receptors, which are expressed late during immune reactions [reviewed in (97, 125)]. Although known as a marker of exhaustion and relaying negative signals to T cells (119), PD-1 plays an important role in the generation of iTregs. Interaction of PD-1/PD-L1 synergizes with TGF-β pathway for induction of iTregs, and mice lacking PD-L1/PD-L2 are impaired for induction of iTregs (60). We have found that Tregs from PD-1-deficient mice are also impaired in their ability to inhibit the activation of innate immune cells (105). However, in the context of cancer biology, PD-1 expression on tumor-infiltrating Tregs is a marker of Treg exhaustion, and PD-1^Hi cells isolated from tumors have lower suppressive activity (147). The TNF receptor family member GITR is also a costimuatory molecule expressed on Tregs (159, 219), and treatment of mice with Fc-GITR-L increased Treg numbers (138). GITR along with the other TNFR family members OX-40, ICOS, and TNFR2 were proposed to regulate the responsiveness of tTreg to IL-2 (189). However, ligation of these receptors also induces proliferation and activation of effector T cells and has been shown to inhibit Treg activity (224).

IL-33/ST2 Pathway

While studying the role of IL-2 in CD4⁺ T cells from the scurfy mice, we found that ST2 (IL1RL1) was one of the most highly responsive cell surface molecules regulated by IL-2 (212). ST2 is the receptor for IL-33, an alarmin expressed in the nucleus of several cell types, which is known to be released on cellular damage (27, 139). Since tTregs are critically dependent on IL-2, they may express ST2 as well. Indeed, we found that FACS-sorted nTregs but not iTregs expressed ST2 mRNA and that a major proportion of Tregs in a naïve mouse expressed ST2 (226). Furthermore, similar to IL-2, IL-33 alone could induce expansion of Tregs systemically. Importantly, IL-2 and IL-33 synergized in vivo to expand Tregs much more potently than either cytokine alone. Similar observations were made by Schiering et al. (201), which showed that ST2-expressing Tregs are enriched in the intestine especially among the Helios⁺ subset. IL-33 treatment promoted proliferation of tTregs and TGFβ-mediated differentiation of iTregs in vitro. They also showed that ST2⁺ Tregs had higher expression of CD103, OX40, and Foxp3. IL-33 treatment in vitro also increased the recruitment of Gata3 and Pol II at the promoter region of the Foxp3 gene. In another recent study, IL-33 treatment increased the production of IL-2 by DC, which contributed to expansion of CD4⁺ST2⁺Foxp3⁺ Tregs (156).

A novel hybrid cytokine (IL233) bearing the activities of IL-2 and IL-33 cytokine domain in a single molecule offered improved synergy to increase Tregs as compared with the co-administration of IL-2 and IL-33 (226). Since IL-33 is released on cellular damage, it may constitute a signal for recruitment of Tregs to prevent inflammation and activation of auto-reactive T cells in response to neoantigens released during cellular damage. Interestingly, helminth infestations induce release of IL-33 from intestinal epithelial cells (85), and helminths have also been shown to increase Treg levels. Helminths also protect from autoimmune Type 1 diabetes and multiple sclerosis in mouse models (55). Clinical trials are ongoing for helminth-induced immunomodulation therapy in relapsing-remitting multiple sclerosis (NCT00645749, NCT01413243). Expansion of Tregs in response to IL-33 may be one of the mechanisms in the therapeutic use of helminths for inflammatory diseases.

TGF-β Pathway

TGF-β is an immunoregulatory cytokine, which signals through a heterodimeric receptor composed of TGF-βRI and TGF-βRII subnits, is a pleiotropic cytokine and inhibits the activation of multiple immune cell types. A role of TGF-β was proposed in Treg biology when it was found that TGF-β1-deficient mice die early with multiorgan inflammatory disease (146). Expression of a dominant-negative form of TGF-βRII or ablation of TGF-βRII on T cells also recapitulated the inflammatory phenotype (154). However, when TGF-βRI was deleted late in antigen-experienced cells, it did not affect Tregs and mice were spared from autoimmunity (146). It was then realized that TGF-β may be dispensable for generation of tTregs but is critical for the induction of pTregs (222). Further studies as discussed below also demonstrated that Smad3, which is downstream to TGF-β signaling, does not bind Foxp3 promoter or the conserved tTreg-specific regulatory region in the Foxp3 gene, rather it binds the pTreg-associated elements (202, 238). Indeed, naïve CD4⁺ T cells (Tn) were shown to be converted to Foxp3⁺ Tregs by TCR stimulation in the presence of TGF-β1 and IL-2 (40). During antigenic stimulation, TGF-β1 can also convert naïve T cells to the regulatory Th3 cells, which do not express Foxp3 but themselves produce TGF-β to regulate immune activation and Foxp3⁺ Tregs (25). Other TGF-β pathway-related molecules have been found to contribute to Treg maintenance and function. These include the latent TGF-β-binding proteins (LTBPs) termed GARP/LRRC32 (44, 239), the integrins αVβ8 (45), and αE(CD103) (72). Interestingly, IL-2 was shown to regulate CD103 expression on helper T cells and contribute to the migration and recruitment of T cells and Tregs to skin as well as gut and lung mucosal surfaces and lung (213).

Adenosine Metabolism and Tregs

Tregs express at least three cell surface receptors that are involved in modulation of extracellular ATP and its role in immune activation that include CD39, CD73, and adenosine 2a receptor (A_2AR) (41). Extracellular ATP released during injury is a potent inflammatory stimulus. CD39 and CD73 expressed on Treg cell surface are involved in dephosphorylation of ATP or ADP to AMP (CD39) and subsequently to adenosine (by CD73). Tregs deficient for CD73 and thus lacking in their ability to produce adenosine were also impaired in their ability to protect in inflammatory disorders (105). By binding to A_2ARs, adenosine inhibits effector T-cell activation, while it also induces differentiation and activation of iTregs (171). A_2AR activation upregulated PD-1 and IL-10 expression on Tregs and promoted their in vivo protective activity in the mainly innate inflammation model of kidney ischemia-reperfusion injury (105).

Other Regulators of Treg Maintenance and Function

Several other regulators of tolerance have now been found to contribute to the generation, maintenance, and function of the various overlapping and nonoverlapping subsets of Tregs. The vast majority of them were initially identified among pTregs in gut-associated lymphoid tissue and have now been extended for other organs. These will be mentioned here, with references provided for further details. The prominent among them are all-trans retinoic acid (ATRA) (15, 38, 227), Vitamin D3 (96), and indoleamine-pyrrole 2,3-dioxygenase (IDO) (207). Exposure of gut and dermal DC to ATRA and Vitamin D3, respectively, increases the expression of IDO, which promote the differentiation and proliferation of Tregs. Short-chain fatty acids (especially butyrate), produced by the gut Clostridium bacteria were also shown to promote differentiation, proliferation, and activation of Tregs (7, 64). The AKT/mTOR pathway is a negative regulator of iTreg differentiation, and the mTOR inhibitor rapamycin increased Treg generation and has been proposed for therapeutic use in inflammatory conditions (75, 198, 270).

TRANSCRIPTIONAL AND EPIGENETIC REGULATION OF TREGs

The realization that Foxp3 plays a central role in Treg lineage and function prompted studies on the regulation of its expression. Although elegant studies have identified several factors for the maintenance and function of Tregs as discussed below, the exact event, which determines Foxp3 expression in thymus is still elusive.

Transcriptional and Epigenetic Regulation of Foxp3

The Foxp3 gene has four distinct regulatory regions (Fig. 2), the promoter region and three conserved noncoding sequence elements (CNS1, CNS2, and CNS3) (277). A unique demethylation pattern is observed at these regulatory regions of freshly isolated Tregs but not on the conventional T cells.

Fig. 2.

Transcriptional regulation of Foxp3. Foxp3 transcription is controlled by binding of a diverse set of transcription factors to distinct regulatory elements in the Foxp3 gene, namely promoter, conserved noncoding sequence (CNS) 1, CNS2, and CNS3. Regulators downstream to TCR/CD28 signaling bind at all the elements. IL-2 signaling recruits activated STAT5 to the promoter and CNS2, while TGF-β-mediated Smad and other peripheral (induced) Tregs related transcriptional factors (RAR, Foxo) bind to CNS1. Foxp3 is an autoregulator and binds to CNS2 as a trinary complex with Cbf-β and RUNX1. Demethylation of CpG islands in CNS2 regulate tTreg stability, and CNS1 may regulate iTreg differentiation, whereas CNS3 may regulate precursors to Tregs. The upstream signals and transcription factors are color coded. Upstream pathways inducing expression of the downstream transcription factors are depicted in the same color. AP1, activator protein 1; c-Rel and p65, members of NF-κB transcription factor family; CREB, cyclic-AMP-responsive element-binding protein; NFAT, nuclear factor of activated T cells; STAT5, signal transducer and activator of transcription; ST2, receptor for IL-33; Gata3, GATA binding protein 3, Pol II, RNA polymerase II; SMAD3, mothers against decapentaplegic homolog; Cbf-β, core-binding factor, β subunit; Foxp3, forkhead box P3; RUNX, runt-related transcription factor 1; Nr4a, nuclear receptor 4a; Foxo, forkhead box O; RAR, retinoic acid receptor. [From Huehn and Beyer (84).]

Promoter.

During T-cell development the CpG motifs in the Foxp3 promoter are completely demethylated in tTregs but are methylated in conventional CD4⁺ T cells, which is even higher on T-cell activation (88, 101, 272). The promoter region gets demethylated during iTreg induction similar to tTregs. Multiple transcription factors that are known to be required for tTreg-cell differentiation bind to the Foxp3 promoter region (reviewed in Ref. 84). These include cAMP-responsive element-binding protein (CREB), nuclear factor of activated T cells (NFAT), activator protein 1 (AP1), signal transducer and activator of transcription 5 (STAT5), nuclear receptor 4a (Nr4a), c-Rel, and p65, and this is consistent with the role of TCR, CD28, and IL-2 signals in the upregulation of Foxp3 expression (Fig. 2). Recently, IL-33 was also shown to induce the recruitment of Gata3 and Pol II to the Foxp3 promoter region and enhance its expression (201).

CNS1.

The CNS1 element, which is in the first intron, does not have any CpG motifs and is regulated primarily by histone modifications (277). Consistent with the role of TGF-β in the differentiation of iTregs and not the tTregs, Smad2/3 only binds in the CNS1 regions, because deletion of CNS1 only affected the differentiation of iTregs but not tTregs (238). Other transcription factors bound to CNS1 region are NFAT and AP1 (277), indicating the requirement of TCR/CD28 stimulation for iTreg differentiation as well as by retinoic acid receptor (RAR) (277), as has been verified by the role of ATRA in pTreg differentiation in the gastrointestinal system (Fig. 2).

CNS2.

The CNS2 region is what really set the tTregs apart from all other Treg and non-Treg T cells, and is a bona fide epigenetic marker for tTreg-lineage. CNS2, also located in the first intron of the Foxp3 gene, is fully demethylated in tTregs and thus is also called Treg-specific demethylated region, TSDR (56, 182, 277). Demethylation status of CpG island in CNS2 is critical for the stability of Treg lineage. This region is methylated in the double-positive thymocytes as well as in single-positive non-Treg CD4⁺ T cells. The TGF-β-mediated differentiation of iTregs also does not induce demethylation of the CNS2 region, eluding to the instability of Foxp3 expression in these cells. Activated CD4⁺ conventional T cells in humans transiently express Foxp3, yet the CNS2 regions in these cells remain methylated, and the cells gradually lose Foxp3 expression (272). Interestingly, deletion of CNS2 region did not induce overt autoimmunity or a complete loss of Foxp3 expression in CNS2-deleted mice, although the frequency of Foxp3⁺ cells and the intensity of Foxp3 expression was reduced as compared with CNS2-sufficient Tregs (52, 135). Methylation of CNS2 is likely an active mechanism to inhibit Foxp3 expression in non-Treg cells as deletion or chemical inhibition of DNA methyltransferase 1 induced expression of Foxp3 on NK and CD8 T cells (93).

The CNS2 region is bound by several transcription factors including Foxp3 itself, indicating the role of Foxp3 and CNS2 in lineage commitment and stability of Foxp3 expression. The other transcription factors bound to CNS2 include c-Rel, p65, NFAT, CREB (TCR/CD28 related), STAT5 (IL-2 signaling), and Runx1-Cbf-β. Runx1-Cbf-β complex binds to demethylated CNS2 region and recruits Foxp3 to the CNS2 element, and deletion of Cbf-β in Tregs led to a loss of Foxp3 expression stability (109, 193).

CNS3.

Studies on chromatin remodeling of the Foxp3 locus shows that the CNS3 regions are the first to open during thymopoiesis and stay open in Tregs as well as all CD4⁺ T cells (277). It is primarily bound by NFκB family transcription factors c-Rel and p65 transcription factors and is believed to mark Treg precursors.

Recently, Foxo1 and Foxo3 were also shown to bind to Foxp3 regulatory regions, and deletion of Foxo proteins in CD4 T cells led to lower proportion and activity of Foxp3⁺ Tregs. Binding sites for Foxo proteins, which are downstream to the AKT/PI-3K pathway, were also discovered in CNS1 and CNS3 regions, but not CNS2, indicating their role in iTregs (98, 173, 174). The complexity of this issue is further illustrated by the finding that deletion of both Foxo1 and Foxo3 by using CD4-Cre resulted in lower proportion and numbers of Foxp3⁺ cells, whereas deletion of Foxo1 by using a Foxp3-Cre resulted in greater numbers of Foxp3⁺ cells; however, their activity was impaired, and in either case the mice were susceptible to multiorgan inflammatory disease (98, 174).

Transcriptional Regulation by Foxp3

Foxp3 itself is a winged-helix DNA binding protein and a unique transcriptional factor that serves both as a transcriptional activator and transcriptional repressor. Deletion of the DNA binding domain of Foxp3 or fusing GFP to the COOH terminus of Foxp3 rendering it inactive helped elucidate the role of Foxp3 as a transcription factor (67, 140). Tregs from these mice retained some Treg-associated programs in terms of CD25, CTLA-4, and GITR expression. These Tregs also retained their anergic status and the inability to produce IL-2. However, now these cells started to express IL-4 and IL-17 under stimulatory conditions (255). Foxp3 can interact with the TH17 transcription factors RORγt and Rorα as well as the interferon regulated factor (IRF)-4 (276, 279). These interactions may be involved in repression of the RORγ and IRF-4-related inflammatory program. In addition, Foxp3 was found to interact with NFAT, AP1, and Runx/Cbf-β, and these interactions might play a role in regulation of the Foxp3-mediated Treg program (see Transcriptional and Epigenetic Regulation of Foxp3 above). Although, Gata3 binds to the Foxp3 promoter, coexpression of Gata3 or T-bet with Foxp3 was shown to enable Tregs to restrict the respective Th2 and Th1 inflammatory responses more efficiently (113, 257).

MicroRNA and Tregs

In addition to the above-mentioned transcriptional programs regulating Foxp3 and regulated by Foxp3, microRNA (miRNA) also have important functions in Treg biology by regulating Foxp3, as well as being regulated by Foxp3. Deletion of two RNA processing enzymes Dicer and Drosha in Tregs induced fatal multiorgan inflammatory disease, similar to the scurfy mice (35, 143). Ablating the miRNA generative capacity of Tregs impaired the peripheral homeostasis and suppressive capacity of Tregs; however, the expression of other Treg-associated markers was maintained (280). Different miRNA species regulate specific components of Treg function, e.g., miRNA-155 controls responsiveness of Tregs to IL-2 (149), whereas miR-142-3p regulates the suppressive function (82). MiR-125a-5p increased the resistance to IL-6-mediated decline of Foxp3 expression (132), and miRNA-146 controls STAT1 signaling in Tregs and their ability to restrict Th1 responses (261). In clinical studies, miR-141 and miR-200a were associated with MS and were proposed to regulate the Treg/Th17 balance (165).

TREG SUPPRESSION MECHANISMS

As discussed above, the suppressor function and lineage differentiation of Tregs are differentially regulated. Several molecular mechanisms have been delineated over the years that are used by Tregs to inhibit activation of the immune system (Fig. 1). Tregs suppress the activation of cells of the innate and adaptive immune system, including T cells, B cells, NK cells, DC, and macrophages, as well as granulocytes. In Foxp3-deficient scurfy mice, T-cell activation can be observed as early as the first week after birth, which rapidly amplifies to almost all the immune cells and production of inflammatory cytokines (210). It is worthwhile to note that in the OTII-transgenic scurfy mice, primarily the T-cell-bearing OVA-specific TCR did not expand and the nontransgenic TCR-bearing cells rapidly expanded and infiltrated the target organs, indicating a specific loss of regulation of autoimmunity (275). Continued expression of TCR was shown to be critical for maintaining the suppressive function of Tregs (131).

IL-2 Consumption

One of the earliest mechanisms proposed for Treg suppressor activity was consumption of IL-2 and outcompeting activated T cells for IL-2, owing to the constitutive high-level expression of CD25 on Tregs (12, 235). In CD25-deficient mice, not only Treg proportions were compromised, IL-2 consumption was also impaired, leading to accumulation of abnormally high circulating IL-2 levels, which induced inflammatory pathology as well abnormal expansion of memory phenotype CD8 T cells (216). Adoptive transfer of CD4⁺CD25⁺ T cells from normal mice normalized the IL-2 and CD8⁺ T cell levels. A recent elegant study showed that IL-2 consumption and downstream STAT5 signaling was important to maintain the suppressor function of Tregs (33). Furthermore, IL-2 capture by Tregs was critical for suppressing CD8⁺ T-cell activation, but was dispensable for controlling the CD4⁺ T cells.

Soluble Mediators

TGF-β1 and IL-10 are immunoregulatory cytokines that are produced by many cell types, including Foxp3⁺ Tregs as well as Tr1 and Th3 cells, and constitute soluble mediators of Treg suppression activity (133, 134, 158, 192). Tregs deficient in these cytokines have impaired ability to regulate immune activation. Other soluble mediators include IL-35, which can induce infectious tolerance by converting naïve CD4⁺ T cells into IL-35-expressing Tr35 cells (36, 37). As mentioned above, Tregs also have the ability to convert pro-inflammatory extracellular ATP to anti-inflammatory adenosine through CD39 and CD73 (41, 105, 171). Adenosine can inhibit the activation of innate and adaptive cells in a paracrine manner and also stimulate Tregs to express PD-1 in an autocrine manner through A_2AR signaling.

Contact-Dependent Mechanisms

An early mechanism proposed for Treg-mediated suppression was identified by the common appearance of CTLA-4 in various genetic screens for autoimmune diseases. Deletion of CTLA-4 impaired the suppressive function of Tregs, without affecting the Treg differentiation (11, 278). CTLA-4 ligation to CD80/CD86 on APCs is critical for maintaining immune tolerance, and the lack of CTLA-4 resulted in higher expression of CD80 and CD86, which provided greater costimulatory signals to the effector T cells (278). We believe that during homeostasis, the MHC II molecules are occupied by self-antigen, prompting Tregs to stay in association with DC and lowering their activation status through CTLA-4 and CD80/CD86 interactions (209). During infections, the self-antigen on MHC II will be replaced with the foreign peptide, which may cause Tregs to dislodge. Furthermore, signaling through DAMPs will upregulate MHC II for display of the pathogen’s antigens on their surface and CD80/CD86 for enhanced costimulatory potential. This will allow the immune reaction to go forward to eliminate the pathogen. However, once the danger signal is gone and the pathogen is cleared, the MHC II of DC will be again loaded with self-antigen, thus allowing Tregs to resume the association with DC for enforcing self-tolerance. Other costimulatory molecules that are used by Tregs to increase their interactions with DC include Lag3 and TIGIT that play a role in the suppressive function of Tregs (83, 269).

Cytotoxicity

Tregs can also directly kill activated immune cells through granzyme B as a soluble mediator. These studies, originally reported in murine models, were also demonstrated using human adaptive Tregs that can induce cytolysis by using granzyme B (69, 273). Another cytolytic mechanism employed by Tregs to kill CD8⁺, but not CD4⁺ T cells, is through Fas-FasL in a cell-cell contact-dependent manner (225). The latter studies were performed in vitro, but do prompt revisiting the relevance of autoinflammatory disease in Fas or FasL mutant mice.

THERAPEUTIC IMPLICATIONS OF TREGs IN INFLAMMATORY RENAL DISEASE

In most, if not all, studies on inflammatory diseases, including acute and chronic renal disease, depletion of Tregs worsens the disease, whereas Treg supplementation offers robust protection (43, 108, 226). Inflammatory diseases of the kidney are a major medical burden. Kidneys are heavily perfused organs and responsible for filtration of blood. The differential distribution of blood flow and partial pressure of oxygen between the renal cortex (50 mmHg) and medulla (10–20 mmHg) makes the medulla especially susceptible to hypoxia (20, 50, 73). Kidneys get up to 25% of cardiac output and constitute only 1% of the body mass, leading to renal accumulation of substances that interfere with cellular metabolism, making them vulnerable (180, 243). Renal manifestations are often secondary to other ailments and contribute to mortality and morbidity. Diabetic kidney disease, either due to Type 1 or Type 2 diabetes, is a major cause of morbidity in the United States, with the incidence on the rise (43). Autoimmune disease such as lupus (SLE), which is marked with high levels of circulating autoantibodies, manifests in half of patients as immune complex deposits in the kidneys and renal failure in up to 10% of patients (5). A sizable number of patients undergoing chemotherapy suffer from nephrotoxicity-associated renal damage, which adds not only to the health burden, but also limits the administration of potent chemotherapeutic drugs. Several other renal disorders such as IgA nephropathy, membranous glomerulopathy, focal segmental glomerulonephritis, etc., are worsened by the immune system. A large proportion of these patients never recover and proceed to end-stage renal disease (ESRD), with transplantation as the only resort. Renal transplantation along with alloreactivity is also plagued by ischemia-reperfusion injury (IRI), which is further amplified by the inflammatory immune response. Last but not least, acute kidney injury (AKI) occurs in a larger proportion of all hospitalized patients and up to 50% of ICU patients. Although a sizable proportion of these patients recover, 5–20% of them progress to ESRD (26). Inflammation plays a major role in the pathogenesis. Thus strategies that can intervene with inflammation should prove to be beneficial. Tregs have proven to be one of the most potent enforcers of peripheral tolerance and can prevent undue activation of the immune system. Below we will discuss the role of Tregs in the major inflammatory renal diseases with the exception of transplantation, the literature on which is vast and deserves a separate review article by itself.

Ischemic Kidney Injury and Tregs

Prevention of AKI by Treg supplementation.

Inflammation plays a major role in the pathogenesis of AKI (9, 87). To test the role of Tregs in AKI, we targeted Tregs by using an antibody against CD25, which depleted ~50% of Tregs and performed IRI (107). Depletion of Tregs before IRI worsened the inflammation, acute tubular necrosis, and renal function assessed by plasma creatinine. In another approach, Rag1-deficient (knockout) mice reconstituted with lymph node cells from Treg-deficient scurfy mice and subjected to IRI had worse renal function, injury, and inflammation as compared with Rag1 knockout (KO) mice that received lymph node cells from Treg-replete wild-type mice (107). Importantly, in the same study, supplementation of scurfy mouse lymph node cells with CD4⁺CD25⁺ cells from wild-type mice inhibited the greater AKI associated with scurfy lymph node cell transfer to Rag1 KO mice. Adoptive transfer of freshly isolated wild-type Tregs to lymphocyte-deficient Rag1 KO mice directly demonstrated that the protective effect of Tregs in kidney IRI does not require adaptive immune cells. Expression of the immunosuppressive cytokine IL-10 was critical for the Treg-mediated protection as adoptive transfer of Tregs from IL-10-deficient mice was unable to protect in renal IRI (107). Mechanistically, Tregs effectively prevent the accumulation of neutrophils and mononuclear phagocytic cells in the kidney during reperfusion, by pathways that are not completely understood to date. Interestingly, Treg depletion with PC61 in a cecal ligation puncture of sepsis-induced AKI improved survival and renal outcomes, suggesting that Tregs protect mainly from sterile renal inflammation (129).

The use of Tregs or Treg-enhancing strategies have been investigated to evaluate protection after the onset of injury in AKI models. Gandolfo et al. (65) found increased trafficking of Tregs in the kidneys when measured 3 and 10 days post-IRI. Depletion of Tregs with an anti-CD25 (PC61) antibody starting 1 day after IRI resulted in continued injury and inflammation in the kidneys, leading to worse renal function and mortality. Adoptive transfer of Tregs 1 day after IRI surgery reduced Th1 cells as measured at day 3 and improved recovery measured day 10 post-IRI.

Preconditioning.

Ischemic preconditioning has been shown to protect from IRI and is believed to be mediated, in part, through increased recruitment of Tregs to the postischemic kidneys (106). Indeed, depletion of Tregs with PC61 after the preconditioning ischemia and before the second ischemic episode reduced the preconditioning-mediated protection from IRI. In another study, reconstitution with Tregs from donor preconditioned mice restored the protection; however, the efficacy of the preconditioned Tregs to restore the protection was not compared with those isolated from nonpreconditioned mice (34). Galectin-9 is a β-galactoside binding protein, which is a urate transporter in the kidneys, is also expressed in myeloid cells, and is known for its role in dendritic cell maturation and inducing apoptosis of promoting iTreg differentiation (151). Combination of galectin-9 pretreatment with ischemic preconditioning offered better protection from IRI than ischemic preconditioning alone (271). The protection correlated with a decrease in IFNγ-producing CD4⁺ T cells and an increase in the proportion of Foxp3⁺ Tregs in peripheral blood, spleen, and kidneys. Galectin-9 treatment by itself without ischemic preconditioning, however, did not protect from IRI.

Liang et al. (137) also investigated preconditioning in the setting of IRI, demonstrating that indeed preconditioning-mediated protection form IRI involves Tregs, as antibody depletion of Tregs took away the protection. In the same study, administration of a superagonistic antibody against CD28 before IRI was also shown to increase Tregs and protect from IRI, with the protection being reversed on Treg depletion. However, there is a caution associated with the CD28 superagonists, because a clinical trial employing TGN1412, a humanized CD28 superagonist antibody, had severe deleterious consequences at a dose 500 times lower than what was considered safe in animal studies (78, 86).

Heat shock preconditioning upregulated heat shock protein 70 (Hsp70) in mice along with an increase in the Tregs (102). Adoptive transfer of T cells from heat-preconditioned mice protected T-cell-deficient nu/nu mice from renal IRI. This protection was lost if Tregs were depleted before the T-cell adoptive transfer and restored when Tregs were replenished. Inhibition of Hsp70 expression by quercetin treatment of the heat-preconditioned mice attenuated the Treg expansion and renoprotection. Furthermore, Tregs from Hsp70-deficient heat-preconditioned mice only offered partial protection in this study.

IL-2 and IL-33.

The role of IL-2 in Treg homeostasis is well established. A complex of IL-2 and anti-IL-2 antibody was shown to induce robust proliferation of Tregs, CD8⁺ T, NK, and NK-T cells (19). Pretreatment of mice with IL-2 and anti-IL-2 complexes expanded spleen and kidney Tregs and offered protection in renal IRI (103). Furthermore, administration of IL-2/anti-IL2 antibody complex 1 day after IRI for 3 days improved the recovery of kidneys with reduced apoptosis and increased proliferation (PCNA⁺) of renal tubular cells (103).

We have found that IL-2 regulates the expression of IL-33 receptor, ST2 (212). The COOH-terminal half of IL-33 resembles the IL-1 family of cytokines and signals through the common IL1RAcP through MyD88/TRAF/NFκB pathway, and was originally thought to be involved with the Th2 response. However, we found that a major proportion of Tregs express ST2 and that Tregs increase in response to IL-33 stimulation, as also shown in other studies (201). Since IL-2 upregulates ST2 and increases the responsiveness of Tregs to IL-33, we treated mice with a mixture of IL-2 and IL-33, which induced a robust and rapid increase in Treg levels systemically and protected mice from IRI much more efficiently than IL-2 or IL-33 administered alone (226). Furthermore, a novel hybrid cytokine (termed IL233), containing IL-2 and IL-33 in a single molecule, was more efficient than the mixture of IL-2 and IL-33 at equimolar ratios to increase Tregs as measured in the spleen and blood. This may likely be due to better targeting of IL-33 to Tregs, owing to using IL-2 as a handle. Pretreatment with the IL233 hybrid cytokine also protected mice from IRI at a one-third lower molar dose as compared with the IL-2 and IL-33 mixture. Importantly, IL233 treatment increased the intrarenal level of Tregs before IRI, which was even more pronounced after renal IRI, indicating the active participation of Tregs in protection from ischemic kidney injury. In addition, IL233 administration after IRI surgery also completely protected mice from loss of renal function and mortality for 28 days when the experiment was terminated, whereas all the control mice were moribund in 3 days. When analyzed at day 28 after the IRI, the IL233 mice were indistinguishable from uninjured mice, without any renal fibrosis or any other structural abnormalities (226).

The spleen is a reservoir for blood as well as immune cells. Our data in this study showed that ischemic injury induces mobilization of Tregs from the spleen, and if the mice were pretreated with IL233, the mobilization was even more robust. Correspondingly, more Tregs were present in the kidneys of IL233-treated mice postischemia as compared with saline-treated controls. CD4⁺CD25⁺ Tregs isolated from IL233-treated mice not only had greater ability to suppress proliferation of conventional T cells in vitro, but they were also able to prevent IRI more efficiently than Tregs isolated from vehicle (saline)-treated mice. In agreement with the improved mobilization of Tregs by IL233 treatment, the Tregs isolated from IL233-treated mice trafficked to the kidneys in greater numbers, whereas the Tregs from the saline-treated mice accumulated at greater numbers in the spleen (M. E. Stremska, V. Sabapathy, and R. Sharma, unpublished observations). IL233 also increases the number of group 2 innate lymphoid cells, ILC2 in the spleen and kidneys, and adoptive transfer of ex vivo-expanded ILC2 also protected from IRI (226). We believe that IL-33 being an alarmin is designed by nature to negate the threats posed by cellular damage and serves as a signal for Tregs (and Th2 and ILC2) to restrict Th1-mediated inflammation. While our study was ongoing, others have also found a protective role for IL-33 in intestinal inflammatory diseases by increasing Treg levels (201).

Sphingosine.

Sphingosine 1 phosphate analog FTY720 (fingolimod) has been shown to be beneficial in inflammatory disease (32). In a mouse model of IRI, FTY720 administration post-IRI increased the levels of CD11b⁺CD11c⁺ DC in the spleens and kidneys, which coincided with a greater expression of Foxp3 mRNA in these organs, suggesting that Tregs may also participate in the FTY720-mediated protection from IRI (104). Pretreament with another natural sphingosine analog sphingosine N,N-dimethylsphingosine (DMS) also protected mice from bilateral IRI with the protection correlating with greater infiltration of total CD4⁺ and Foxp3⁺ cell in the kidneys before IRI (120). Inhibiting Tregs by blocking CD25 or CTLA-4 negated the DMS mediated protection. However, in this study DMS administration after IRI did not offer any protection.

Adenosine and other nucleotides.

Interactions of adenosine with the A_2AR offer protection from inflammatory disease, and Tregs are involved in generation of adenosine through cell surface expression of CD39 and CD73 (41). An autocrine role of the adenosine pathway was identified in Tregs that contributes to the protective function of Tregs in IRI (105). Adoptive transfer of CD73 or A_2AR-deficient Tregs was unable to protect in IRI model, indicating an intrinsic role of adenosine generation and utilization in Treg protective ability. Treatment of CD73 KO Tregs with A_2AR agonist ATL1222 before the adoptive transfer restored their protective potential. Furthermore, ex vivo pharmacological stimulation of A_2AR on Tregs promoted their protective ability such that as little as 40,000 Tregs significantly inhibited IRI-associated renal dysfunction and injury.

Extracellular ATP is considered pro-inflammatory through binding to the purinergic P2X and P2Y receptors and blocking of P2X7 receptor, which is highly expressed in the kidneys was beneficial in inflammatory renal conditions (28). In a recent study, Koo et al. (114) investigated the use of periodate-oxidized ATP (oATP), a P2X7 receptor (P2X7R) antagonist (121), in the settings of IRI. Mice treated with oATP for 7 consecutive days until the day of IRI were protected from renal dysfunction, tubular injury, and inflammation (114). Accordingly, the P2X7-deficient mice were protected from IRI. Interestingly, this protection was associated with an increase in Tregs in the kidneys, and the protection was abrogated on Treg depletion. Furthermore, adoptive transfer of P2X7-sufficient, but not P2X7-deficient, Tregs showed protective response to oATP treatment in immunodeficient Rag1 KO mice from IRI; however, the P2X7 KO Tregs were more protective than the WT Tregs (114). Treatment of mice with oATP also protected mice from renal dysfunction and fibrosis when administered after IRI (114).

PD-1 and PD-1 ligands.

Microarray comparison of Tregs from A_2AR-sufficient and deficient Tregs showed that PD-1 expression on Tregs is linked with A_2AR stimulation (105). The protection offered by Treg-A_2AR stimulation was abrogated if the Tregs were treated with a blocking antibody to PD-1. Subsequently, Jaworska et al. (89) showed that the expression of PD-1 on Tregs and its ligands PD-L1 and PD-L2 in recipient mice are important for protection from renal IRI.

Other anti-inflammatory mechanisms in IRI.

Mesenchymal stem cells (MSC) have long been proposed to have anti-inflammatory properties in addition to their tissue regenerative capacity. The efficacy of MSC to intervene in renal IRI has been investigated (81). MSC infused 6 h after IRI surgery protected from renal inflammation and injury. Importantly, a greater number of Tregs accumulated in the spleen and kidneys of MSC-infused mice. This protection was abrogated if the mice were splenectomized or if Tregs were depleted using an anti-CD25 antibody.

Mycophenolate mofetil (MMF) is often used post-renal transplant to prevent inflammation, and it was found that in an animal model of IRI, MMF could alter the recruitment of Tregs to the kidneys along with CD4⁺ and CD8⁺ T cells and may affect the contribution of Tregs to renal repair (66).

Nephrotoxic Renal Injury and Tregs

Different classes of nephrotoxins induce diverse renal pathology in humans and mice. The most widely studied nephrotoxic injuries include cisplatin, Adriamycin (doxorubicin), gentamycin, aristolochic acid, and folic acid. However, the role of Tregs has mostly been studied in the former two models only. Whereas cisplatin primarily induces tubular damage, Adriamycin can induce both tubular and glomerular injury and is often proposed as a model of focal segmental glomerulosclerosis (130). Irrespective of the involvement of renal segments, inflammation plays a major role in the amplification of the injury and pathogenesis. The role of Tregs in preventing inflammation and injury and therapeutic use of Tregs or Treg-enhancing strategies have been explored in renal manifestations of these conditions as discussed below.

Tregs in cisplatin-induced renal injury.

Lee et al. (127) were the first to demonstrate that CD4⁺CD25⁺ Tregs protect from cisplatin-induced renal dysfunction and damage. Depletion of Tregs in WT mice exacerbated the cisplatin-induced injury, whereas adoptive transfer of Tregs either in the T-cell deficient nu/nu mice or intact Balb/c mice reduced the serum creatinine and BUN levels along with reduced renal damage, reduced renal levels of TNFα and IL-1β, as well as reduced mortality associated with cisplatin. Kim et al. (99) first showed that bee venom increased Treg proportions and protects from cisplatin-induced renal injury and later demonstrated that a constituent of bee venom, phospholipase A2 (PLA2), modulated Treg numbers and protected mice in a model of cisplatin-induced AKI (100). PLA2 increased Treg proportions in vitro and also induced greater accumulation of IL-10 and IL-10-producing Tregs and DC in the kidneys after cisplatin treatment. The PLA2-mediated renoprotection was abolished after anti-CD25 antibody depletion of Tregs. Importantly, bee venom-associated increase in Tregs and renoprotection did not affect the antitumor activity in tumor-bearing mice. Among other regulators of Treg function in prevention of AKI, it was recently shown that TLR9-deficient Tregs were not able to inhibit cisplatin-induced AKI, although the level and in vitro suppressive function of Tregs in TLR9 KO mice was intact (3). The authors postulated reduced expression of adhesion molecules CD11a and CD44 on Tregs from TLR9-KO mice to be responsible for the lack of protective ability.

As mentioned above, MSCs have protective properties in the renal IRI model (81). Recently, the effect of MSC administration was evaluated in cisplatin-induced AKI (177). Treatment with umbilical cord blood-derived MSCs 1 day after cisplatin administration attenuated intrarenal MCP-1 and IL-6 and showed higher expression of IL-10 and VEGF. This was postulated to be due to an increase in Tregs in the kidneys, because depletion of Tregs after MSC treatment abrogated the protection. However, treatment of mice on day 3 after cisplatin did not increase Tregs or protect renal function, indicating that MSCs do not promote renal repair in cisplatin nephrotoxicity (177).

Pretreatment of mice with the IL233 hybrid cytokine, which increases Tregs rapidly, was also able to rescue mice from nephrotoxic injury induced by cisplatin in C57Bl/6J mice, as observed with lower plasma creatinine and BUN levels and preserved renal architecture. This protection was preceded with an increase in circulating Treg levels (226). Interestingly, in a study by Akcay et al. (2), IL-33 was shown to exacerbate injury and inflammation in the cisplatin model and increased recruitment of CD4⁺ T cells to the kidneys in a CXCL1-dependent manner. IL-33 levels were increased in the blood and kidneys following cisplatin administration. Furthermore, coinjection with cisplatin and IL-33 worsened the injury as compared with cisplatin alone. However, the effects of IL-33 treatment on Tregs was not studied. The difference observed between this study and our studies could be that we pretreated mice with IL233 hybrid cytokine before administering cisplatin, and IL-33 was injected as a fusion to IL-2, thus targeting the majority of IL-33 activity to Tregs, which offered protection.

Tregs in Adriamycin nephropathy.

Foxp3 is the lineage-defining transcription factor for Tregs. Retroviral transduction of CD4⁺ T cells by Fontenot et al. (57) and Hori et al. (77) imparted Treg-like phenotype and regulatory function. A similar strategy was adopted by Wang et al. (252) in one of the earliest studies to show that adoptive transfer of T cells with a retroviral transduction with Foxp3 gene, 1 wk after induction of Adriamycin, protected the mice from renal dysfunction and injury. Subsequently, the same group showed that adoptive transfer of Tregs isolated from wild-type BALB/c mice protected syngeneic SCID mice from Adriamycin (152); however, the Treg-mediated protection is dependent on stimulation with self-antigen, as Tregs isolated from DO11.10 TCR-Tg mice, in which most T cells have specificity to ovalbumin, were unable to protect. Furthermore, the Treg-adoptive transfer-mediated protection was attenuated if the mice were treated with a neutralizing antibody to TGF-β. TGF-β and IL-10 induce distinct categories of immunomodulatory macrophages (M2c) compared with M2a macrophages, which are induced by IL-4 and IL-13 (148). M2c macrophages were more potent in inducing Tregs and protecting mice from the glomerulosclerosis, tubular atrophy, interstitial expansion, and proteinuria induced by Adriamycin, as compared with M2a macrophage. The advantage of M2c over M2a was lost after Treg depletion. In line with the protective role of adenosine in protection from AKI (105, 172), CD25⁺ Tregs from mice transgenic for human CD39 (hCD39-Tg) offered greater protection in the Adriamycin model as compared with either wild-type CD25⁺ Tregs or CD25⁻ T cells from hCD39-Tg mice (251).

In our own studies, pretreatment with the IL-2 and IL-33 containing hybrid cytokine (IL233), which offered protection from ischemic and cisplatin injury, also potently protected BALB/c mice from Adriamycin (226). IL233 treatment as late as 2 wk post-Adriamycin infusion reversed ongoing nephropathy and protected kidneys from renal dysfunction, inflammation, injury, and fibrosis, indicating that enhancing Tregs with IL233 late in nephrotoxic injury models may be a viable strategy to limit ongoing injury and promote repair (V. Sabapathy, N. Cheru, M. Saleh, and R. Sharma, unpublished observations).

Tregs in Glomerulopathies

Autoimmune diseases such as systemic lupus erythematosus (SLE or lupus), Goodpasture’s syndrome, IgA nephropathy (idiopathic), and others often have glomerular manifestations. Defects in Foxp3⁺ Tregs have been indicated in these conditions, especially in lupus glomerulonephritis (GN) as discussed below.

Tregs in Lupus Glomerulonephritis

Lupus GN and Treg deficiency.

SLE is a systemic autoimmune disease manifested as increased production of autoantibodies, immune complex deposition, and multiorgan inflammation (39, 240). Kidneys are affected in over 50% of lupus patients, leading to glomerulonephritis (29), with a fifth of these patients proceeding to ESRD (253). Defects in IL-2 production by T cells from lupus patients were found to contribute to autoimmunity by Lippe et al. (142) in patients and mouse models. IL-2 deficiency interferes with Treg homeostasis and activation-induced cell death, and promotes differentiation of T follicular helper (TfH) (10, 90, 168) and TH17 cells (16, 124) (241). Clinical and experimental studies show that Treg deficiency correlates with incidence and severity of lupus (reviewed in Ref. 118). We will review the studies for lupus GN below.

Xing et al. (263) found that in a cohort of 60 lupus patients the proportion of CD4⁺CD25⁺Foxp3⁺ Tregs, as well as the level of Foxp3 expression per cell (measured as the mean fluorescence intensity of Foxp3), was lower in SLE patients as compared with healthy controls. These values were further reduced in patients with active lupus GN. This was accompanied by a corresponding increase of Th17 cells in SLE patients that was further increased in SLE patients with GN, indicating that an imbalance in Treg/Th17 ratios contributed to renal pathogenesis lupus patients. A similar Treg/Th17 imbalance was observed in another study with a cohort of 42 patients with new-onset lupus nephritis (262). Additionally, lower serum levels and higher urinary levels of TGF-β1, also correlated with higher lupus scores and increased nephritis in this cohort, indicating a diagnostic value of urinary TGF-β1 with disease activity.

Another independent study comparing 17 patients with lupus and 20 patients with lupus GN showed greater Th17 and reduced CD4⁺CD25⁺Foxp3⁺ Treg cells in patients with nephritis compared with SLE patients without nephritis (23). A recent study suggested that reduced numbers of regulatory CD4⁺CD25⁺Foxp3⁺ cells as a biomarker of the activity of SLE, particularly of renal involvement. In a pilot histological study, kidney biopsy samples showed lower proportions of Foxp3⁺ cells in patients with active proliferative class IV nephritis as compared with that of that of ANCA glomerulonephritis, acute tubulointerstitial nephritis, and nephroangiosclerosis (1). Coexpression of IL-17 in Foxp3⁺ cells (called Treg17) was proposed to impart Tregs with increased ability to suppress Th17 responses (110). By using a pristane-induced lupus GN model, abrogation of STAT3 expression specifically in Tregs resulted in higher disease score, which correlated with reduced expression of CC chemokine receptor 6 (CCR6) in Tregs, suggesting that STAT3-mediated CCR6 expression may enable Treg17 to prevent Th17-mediated pathogenesis. In contrast, a recent study demonstrated that Tregs coexpressing Foxp3 and the Th17 lineage-specific transcription factor Rorγt inhibited the anti-inflammatory Th2 responses in the pristane-induced lupus nephritis model (112). A Foxp3-creRORCf^l/fl -mediated ablation of IL-17 production in Tregs ameliorated the vasculitis and nephritis phenotypes.

Treg-based therapeutic approaches in Lupus GN.

Using a model of d3-thymectomy-induced accelerated GN in NZM2328 mice, Bagavant et al. (8) showed that adoptive transfer of freshly isolated Tregs prevented production of autoantibodies and inflammation in prostate, thyroid, and lacrimal glands; however, this did not inhibit GN. Scalapino et al. (200) showed that ex vivo-expanded CD4⁺CD25⁺CD62L^Hi cells reduced the incidence and progression of GN in NZB/W F1 mice, suggesting that increasing Tregs may be a viable therapeutic approach for lupus GN.

Other strategies that were found to be beneficial in lupus GN were accompanied with an increase in Tregs. In clinical trials, depletion of B cells with rituximab correlated with increased levels of Tregs (244). Although not demonstrated directly, treatment with abatacept (CTLA4-Ig) in combination with MMF and glucocorticoids improved autoantibody levels, complement and urinary protein levels (62), and was hypothesized to function through increasing Treg abundance (220). In preclinical studies, a tolerogenic regimen with a low-dose infusion of nucleosomal peptide in lupus-prone SNF1 mice was effective at reducing production of inflammatory mediators and restricted leukocyte infiltration in kidneys (95). Importantly, the treatment induced differentiation of CD4⁺CD25⁺ and CD8⁺ Tregs, which produced TGF-β1 for mediating suppression. Adoptive transfer of the Tregs expanded by the nucleosomal peptide also suppressed lupus-associated autoimmune responses in recipients. Long-term oral treatment of lupus-prone NZB/WF1 mice with curcumin attenuated production of autoantibodies, immune complex deposition, renal inflammation, and proteinuria (126). This protection was abrogated by treatment with PC61, suggesting that the interaction of curcumin with Tregs participated in the protection.

IL-2 treatment to expand Tregs for treatment of autoimmune diseases has been investigated experimentally (70, 128, 145). Low-dose IL-2 therapy improved the outcomes in Graft versus host disease (GVHD) and Hepatitis C virus (HCV) associated vasculitis (155, 194). Treatment with low-dose IL-2, although beneficial, may also induce deleterious effects at high concentrations (13, 49, 175). In two recent clinical trial, low-dose IL-2 therapy corrected T-cell defects by expanding Tregs, inhibiting TfH and TH17 cells, and also lowered the SELENA-SLEDAI disease index; however, data for GN was lacking (76, 246). Mizui et al. (162) showed that treatment of MRL/lpr mice with recombinant AAV-expressing IL-2 increased Tregs and reduced inflammation. However, this effect was attributed to an increase in CD8⁺ T cells and NK cells. In another experimental study, treatment of NZB/WF1 mice with an IL-2/anti-IL-2 antibody complex increased Tregs in spleen and kidneys and reduced autoantibody levels (267). This was accompanied with reduced renal immune complex deposition, lower inflammation in the kidneys, lower proteinuria scores, and decreased mortality. Importantly, the treatment was more effective than the MMF and steroid combination therapy.

Lupus-prone NZM2328 mice have a kinetic loss in the ability to produce IL-2, which led to an age-dependent decrease in total Tregs, as well as ST2-expressing Tregs (R. Sharma, M. Stremska, C. Dai, S. Mohammad, S. M. Fu, unpublished observations). A synergistic treatment with IL-2 and IL-33, especially in the form of IL233 hybrid cytokine, which bears the activity of IL-2 and IL-33 in a single molecule and was found to protect mice from AKI (see IL-2 and IL-33 above), also inhibited the onset of lupus GN in an accelerated model using NZM2328 mice. A one-time induction therapy with daily injection for 5 days, especially with the IL233 hybrid cytokine, increased Tregs in the renal lymph nodes for several weeks and inhibited the production of IFNγ and TNFα. Importantly, IL233 administration after the onset of proteinuria, induced persistent remission lasting for months in >85% of mice. IL233 treatment also protected MRL/lpr mice from proteinuria and mortality accompanied by an increase in Tregs and attenuation of pro-inflammatory cytokines.

Tregs and Nephrotoxic Serum Nephritis

The nephrotoxic serum nephritis (NTS) model is commonly used to study rapidly progressive glomerulonephritis, which culminates in glomerulosclerosis (94). It is induced by injection of antibodies raised either in rabbits or sheep that are directed against the renal glomerular basement membrane (GBM). The disease is characterized by rapid inflammation and infiltration of leukocytes in the kidneys. In the very first report (258) for the therapeutic use of Tregs for GN, adoptive transfer of 1 million CD4⁺CD25⁺ Tregs before induction of anti-GBM GN reduced inflammation and protected from renal damage. By using eGFP marked Tregs, it was found that the transferred Tregs trafficked to the lymph nodes. This trafficking was important for the protection because in another independent study Tregs deficient for CCR7 were unable to protect mice (47). In this model, IL-9 production by Tregs was important for recruitment of mast cells to the lymph nodes, which was required for the protection, as adoptive transfer of Tregs from IL-9-deficient mice did not protect from NTS, and mast cell-deficient mice were also found to be more susceptible, thus indicating a yet unidentified role of a Treg and mast cell nexus in protection from GN (48). Although not shown, mast cells or other myeloid cells may participate in providing costimulation for Treg-mediated protection because in an unrelated study treatment of rats with a super-agonistic antibody to CD28 (JJ316) robustly increased Tregs and protected rats from NTS (230). A heightened Th1 response due to Treg depletion plays a pathogenic role in NTS (179), and increased expression of the Th1-specific transcription factor T-Bet in Tregs made them better suppressors of Th1-driven GN, partly by improving the trafficking of Tregs to Th1-target regions in a CXCR3-dependent manner (167). However, Tregs coexpressing the Th17-specific factor RORγt contributed to NTS pathogenicity (111).

Studies with adoptive transfer of Tregs, although promising, should be carried out carefully, because during polyclonal expansion of Tregs using anti-CD3/CD28 stimulation with TGFβ and IL-2, there is a possibility of generating inducible Tregs, which in a recent study were shown to be ineffective in protecting from experimental sheep anti-GBM-induced GN and failed to suppress Th1 responses (68).

Tregs and Diabetic Kidney Disease

Diabetic kidney disease has become the largest contributor to ESRD and mortality (176, 266). Defects in IL-2 and Tregs are much appreciated in autoimmune T1D (183, 206), with clinical trials ongoing for Treg-adoptive therapy and Treg enhancement strategies (18). Recent studies show that inflammation contributes to ESRD in Type 2 diabetes (T2D) also (51, 115, 163), which is primarily considered a metabolic syndrome. In a clinical study, compared with healthy controls, circulating levels of Tregs were significantly lower in T2D patients with ongoing micro- and macroalbuminuria, and significant inverse correlations were noted between the disease course (measured as urinary albumin excretion rate) and circulating Treg levels (264). In experimental studies, Treg depletion with anti-CD25 antibody worsened proteinuria in diabetic mice, whereas adoptive transfer of Tregs was found to be protective (46). Treatment of genetically obese diabetic mice with the IL-2 and IL-33-containing IL233 hybrid cytokine (similar to the AKI and lupus GN studies) rapidly and robustly increased Foxp3⁺ Treg in lymphoid organs and adipose tissue, accompanied with greater accumulation of M2 macrophages. Importantly, the IL233 cytokine treatment induced normoglycemia, improved glucose tolerance, and protected mice from renal dysfunction (R. Sharma, S. Mohammad, M. Stremska, unpublished observations). In addition, the IL233 treatment also increased the recruitment of ILC2 to the adipose tissue, which may participate in regulating the metabolic syndrome, indicating that Treg modulation by IL-2 and IL-33 can be an attractive strategy to not only restrict inflammation but also regulate metabolism in a Treg-dependent manner. Indeed, the discovery that metabolism of Tregs is different—in terms of preference toward fatty acid oxidation as a source of energy as compared with preferred glycolytic mechanisms by inflammatory T cells—has evolved our thinking in the interplay of Tregs and metabolism (161).

CONCLUSION AND FUTURE IMPLICATIONS

The immune system is designed to constantly patrol the body to look for pathogens and other danger signals. We have begun to appreciate the role of the immune system in regulation of not only threats from pathogens but also regulation of cues that are macro- or microenvironmental. Although several mechanisms have been proposed for loss of tolerance during insult and injury, the onset of autoimmunity is still a black box. Nevertheless, Tregs play an important role, as evidenced above in several inflammatory diseases, including those in the kidney. In addition to their anti-inflammatory potential, Tregs may also serve as an indicator of successful immunomodulatory therapy, for example, in patients with membranous nephropathy, where at baseline Tregs were low and on rituximab therapy the Treg levels increased and were indicative of improvement of disease score, whereas in the nonresponders Treg levels were unchanged as compared with pretreatment baseline (190). Multiple anti-inflammatory mechanisms have been identified for renal and other inflammatory diseases, all of which converge on the central role of Tregs. Inflammatory pathways have the ability to adapt and change their course if one pathway is restricted. Therefore, multiple pathways need to be simultaneously evoked to offer optimal protection in these conditions. In our recent studies mentioned above, use of two pleiotropic pathways, an adaptive (IL-2) and an innate (IL-33), have shown remarkable synergy in strengthening the survival and function of Tregs and other renoprotective cell types, including group 2 innate lymphoid cells (188, 250). It is important to employ the benefits of IL-33 to our advantage, as has been demonstrated by multiple recent studies. By itself, IL-33 may have deleterious effects (42, 136), but directing IL-33 activity preferentially to Tregs, with the help of IL-2, may be a promising clinical approach to therapeutic Treg enhancement (Fig. 3). IL-33 and other mediators of cellular damage such as adenosine, sphingosine, and heat shock proteins are natural signals of cellular damage, which have the potential to activate inflammatory immune responses through release of intracellular DAMPs and neoantigens. These signals, therefore, also serve to activate Tregs to simultaneously limit collateral damage due to a runaway inflammatory response. Although techniques for manufacturing Tregs and Treg-adoptive therapy have progressed significantly with clinical trials ongoing in multiple disorders, many challenges remain, and strategies to improve endogenous Tregs are likely to offer significant advantages. One caveat is in conditions where endogenous Tregs are genetically defective, such as IPEX, where a combined approach of Treg-adoptive transfer and Treg-enhancing strategies may need to be employed for best protection.

Fig. 3.

Treg-enhancing strategies to protect from renal inflammatory diseases. In addition to the adoptive transfer, several other approaches have been used to promote the number and function of endogenous Tregs. These include small molecules such as fingolimod (FTY720), mycophenolate mofetil (MMF), adenosine receptor (A₂AR) agonists, oxidized ATP (oATP), constituent of bee venom (PLA2); cell transfusions such as mesenchymal stem cells (MSCs), tolerogenic dendritic cells (DC); macromolecules including, rituximab, CD28 superagonists, low-dose IL-2, IL-2/anti-IL-2 complex and the recently described IL233 fusion cytokine, which bears IL-2 and IL-33 activities in one molecule. IL-2 upregulates the ST2 (IL-33 receptor) on Tregs enabling IL-2 and IL-33 to synergize and promote activation, proliferation, and recruitment of Tregs to protect from renal inflammatory diseases induced by ischemia-reperfusion, nephrotoxicity, lupus nephritis, and diabetic nephropathy, all of which contribute to end-stage renal disease (ESRD). Please see the text for details.

GRANTS

Support for this study was provided by National Institutes of Health Grants 1R01-DK-105833 [to R. Sharma (contact PI) and Shu Man Fu], 2R01-AI-116725 (subcontract from Fred Perrino, Wake Forest University, to R. Sharma), 1R01-DK-104963 (to G. R. Kinsey), 1R21- DK-112105 (to G. R. Kinsey), a UVA AstraZeneca Research Alliance award (to R. Sharma), and the LaunchPad Diabetes Fund (to R. Sharma). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health or other funding agencies.

DISCLOSURES

A patent (US9840545 B2) for IL233 hybrid cytokine was awarded on December 12, 2017.

AUTHOR CONTRIBUTIONS

R.S. conceived and designed research; R.S. performed experiments; R.S. analyzed data; R.S. and G.R.K. interpreted results of experiments; R.S. and G.R.K. prepared figures; R.S. and G.R.K. drafted manuscript; R.S. and G.R.K. edited and revised manuscript; R.S. and G.R.K. approved final version of manuscript.