The alarmingly high incidence of obesity leading to diabetes and diabetic nephropathy (DN) is the largest contributor to chronic kidney disease and end-stage renal disease (ESRD) in the modern world (Collins et al., 2012). The current treatment strategies including better glycemic control and inhibitors of the renin-angiotensin system, are unable to completely restrict the DN-associated ESRD (Moreno et al., 2018). A growing collection of evidence points to inflammation as a key factor in the pathogenesis of DN (Pichler et al., 2017). Cellular metabolism and hemodynamic changes within the kidney trigger the activation of inflammatory pathways, causing functional and structural renal injury (Wada and Makino, 2013). Therefore, there is a potential in targeting the therapies to improve glycemic and hemodynamic control, as well as to reduce the inflammation and prevent renal damage. Recent studies show protective role of the anti-inflammatory T-regulatory cells (Tregs) in DN (Eller et al., 2011). Micro- and macro-albuminuria in diabetic patients had a significant inverse correlation with the levels of Tregs (Xu et al., 2009). Similarly, antibody-mediated depletion of Tregs worsened the proteinuria in diabetic (db/db) mice, while Treg-supplementation was protective (Eller et al., 2011).

Recent clinical studies show that an imbalance in the ratios of Tregs and Th2 cells to Th1 and Th17 cells contribute to overt DN in T2D patients and regulates the progression of T2D itself (Wu et al., 2010; Zeng et al., 2012). A therapeutic agent that can promote Tregs and Th2 cells, and restore the balance will be beneficial in blocking inflammation in DN and ESRD. We have identified the cooperativity of the signaling pathways of interleukin (IL)-2 and IL-33 (Stremska et al., 2017; Sharma and Kinsey, 2018). The role of IL-2 in promoting Tregs and immune-tolerance is well established (Fontenot et al., 2005; Bayer et al., 2007). IL-33 is an alarmin and was initially perceived as a proinflammatory cytokine (Liew, 2012). However, recent studies have demonstrated its role in Treg-biology and suppression of sterile inflammation (Peine et al., 2016). We hypothesized that owing to the expression of both IL-2 and IL-33 receptors, their cooperativity will promote the homeostasis of Tregs, Th2, and ILC2 and inhibit inflammation. To this effect, we generated a novel hybrid cytokine (IL233) bearing the activities of IL-2 and IL-33 in a single molecule, with a perceived potential to suppress multiple pathways of inflammation (Stremska et al., 2017). Indeed, IL233 outperformed IL-2 and IL-33 given either alone or in combination and protected mice from ischemic as well as nephrotoxic renal injury (Stremska et al., 2017). In the current study, we investigated the hypothesis that a therapy with the IL233 hybrid cytokine can be utilized to suppress inflammation associated with DN to prevent the progression to ESRD in a mouse model of obesity-linked DN.

Diabetic nephropathy (DN) is a chronic renal disease, which has become one of the leading causes of ESRD and mortality globally (Collins et al., 2012). Although DN also occurs in association with autoimmune type-1 diabetes, obesity and obesity-linked diabetes is the major contributor to DN, and is associated with more than 60% cases of ESRD (Wada and Makino, 2013). Hyperglycemia and the downstream factors that are believed to drive pathogenesis include: generation of advanced glycation end products (AGE), protein kinase C-mediated generation of reactive oxygen species (ROS) or mechanical stress/shear injury due to renal hyperfiltration (Wada and Makino, 2013; Pichler et al., 2017), all of which can induce inflammation that plays a major role in the etiopathogenesis of DN.

Current treatment regimen for DN is directed toward optimizing glycemic control and lowering cardiovascular risk. However, recent studies have shown that most successful treatment strategies also contain immune-targeting properties (Moreno et al., 2018). Data from both animal and clinical studies have demonstrated a positive correlation of increase in inflammatory mediators with the pathogenesis (Pichler et al., 2017). Greater expression of chemotactic proteins and adhesion molecules in the kidneys, and a related accumulation of pro-inflammatory M1 macrophages has been reported in DN (Chow et al., 2004). Genetic ablation of macrophage migration associated molecules such as CCR2 or its ligand MCP1 attenuated diabetic kidney disease (Chow et al., 2007). Depletion of macrophages or inhibiting the survival of monocyte-derived macrophages was also shown to be protective in DN (Lim et al., 2009). Although, no direct role of T-cells has been established in the pathogenesis, higher numbers of circulating T-cells have been reported with heightened severity of DN (Bending et al., 1988; Moon et al., 2012). Accordingly, the T and B-cell deficient Rag1-deficient mice were protected from diabetic kidney disease as compared to Rag1-sufficient mice (Lim et al., 2010).

A higher ratio of Tregs to Th1 or Th17 cells was found to be protective against T2D and its downstream vascular complications (Zeng et al., 2012). Further, the ratio of Treg to Th1 or Th17 was much reduced in T2D patients as compared to healthy controls and was even more pronounced in patients with microvascular complications. In diabetic patients with overt nephropathy, the ratio of the Th1 to Th2 cytokines was drastically altered and Th1 cytokines were detectable in the urine of diabetic patients with overt nephropathy (Wu et al., 2010). Our data shows that in the BTBR^ob/ob mouse model, the homozygous Ob mice had fewer Tregs than Het mice and IL233 treatment restored these levels, indicating a correlative role of Tregs with diabetes and DN (Supplementary Figure 3). Compared to the control mice, treatment with IL233 showed higher Foxp3 expression in the kidneys with no significant difference in Tbx21 or Gata3 levels. Systemic low-grade inflammation may be a contributing factor for the lower levels of Tregs in obesity, however, it is interesting to note that leptin has been shown to induce proinflammatory cytokines and inhibit Tregs in leptin/leptin receptor – sufficient mice fed high-fat diet (Matarese et al., 2010; Procaccini et al., 2010). Yet, the leptin-deficient BTBR^ob/ob mice had heightened inflammation and reduced Treg proportions, indicating that factors other than leptin pathway may contributes to lower Tregs. This may be dictated by heightened innate immune cytokines in response to advanced glycation end-products, activation of inflammasomes, metabolic stress and TLR signaling (Akasheh et al., 2013; Gonzalez et al., 2013; Wada and Makino, 2016). Indeed, in our study IL233 treatment lowered inflammation, reduced hyperglycemia, and glycation index, whether administered either before or after the onset of diabetes and microalbuminuria.

Interestingly, Tregs have been shown to favor oxidative phosphorylation and fatty acid oxidation (Michalek et al., 2011; Priyadharshini et al., 2018) for metabolism, yet the hyperlipidemic animals in our current study and T2D patients (Wagner et al., 2013) had reduced Treg levels. This raises several important conundrums that need to be addressed experimentally in future studies. In animal studies, depletion of Tregs with anti-CD25 antibody accelerated DN in db/db mice and adoptive transfer of Tregs attenuated proteinuria and glomerular hypertrophy (Eller et al., 2011). Thus, Tregs offer an attractive approach for treatment of DN. However, generating and use of ex vivo expanded autologous or allogeneic Tregs has technical and biological limitations (Trzonkowski et al., 2015). Therefore, strategies that enhance endogenous Tregs are likely to be beneficial for therapy of DN.

Recently, we demonstrated a role for IL-33 in promoting Treg responses, such that IL-2 and IL-33 can synergize to increase Tregs and protect from experimental acute kidney injury (Stremska et al., 2017). Accordingly, we designed a hybrid cytokine (termed IL233) bearing the activities of IL-2 and IL-33 in a single molecule, which outperformed the mixture of IL-2 and IL-33 at equimolar ratios (Stremska et al., 2017). As anticipated IL233 treatment enhanced Tregs rapidly and in a sustained manner in the Ob mice, which correlated with strong inhibition of diabetes and DN. IL233 treatment also increased the ratio of Tregs to TNFα and IFNγ (Th1) cells in the Ob mice (Supplementary Figure 4). Higher TNFα levels correlate with increased severity of renal injury in diabetic patients (Navarro et al., 2006) and anti-TNF treatment attenuated proteinuria in experimental studies (Moriwaki et al., 2007). Interestingly, in our other recently published studies in Adriamycin nephropathy, intervention with IL233, 2 week after the induction of injury not only inhibited inflammation and restored renal function, but we also observed an induction of a reparative program marked by high expression of renal progenitors in the kidneys of mice treated with IL233 (Sabapathy et al., 2019). This was accompanied by an increase in the Tregs and ILC2 in the kidneys. Further, administration of an antibody to CD25, which depleted both Tregs and ILC2, ameliorated the protective effects. It is possible that a similar reparative response may also contribute to the observed protection from DN in the Ob mice.

The role of Tregs in type 1 diabetes is well appreciated and clinical studies for Treg-therapy have proven beneficial (Yu et al., 2018). Considering that defects in IL-2 production contribute to T1D, low-dose IL-2 therapy has been used to enhance Tregs experimentally as well as clinically with mixed results (Dwyer et al., 2016). Another autoimmune ailment that can lead to ESRD is systemic lupus erythematosus (SLE), where defects in IL-2 production and Tregs have been shown to contribute to pathogenesis (Mizui and Tsokos, 2016), and IL-2-based approaches to enhance Tregs have been shown to be beneficial for SLE but not for lupus nephritis (He et al., 2016; von Spee-Mayer et al., 2016). The use of IL-2 based approaches have been investigated for acute renal diseases including ischemia reperfusion and nephrotoxic injury (Kim et al., 2013; Stremska et al., 2017). IL-33 is an alarmin cytokine expressed in almost all cell types that is released from damaged or dying cells (Liew, 2012). Although, IL-33 has been proposed as a marker of acute renal injury, IL-33 levels were found to be similar between CKD patients and healthy controls. Soluble ST2 (sST2) can act as a decoy receptor to neutralize IL-33 activity and its levels were found to be higher in the circulation of CKD patients (Bao et al., 2012). Another study found that type-2 diabetes patients had elevated levels of IL-33 in the blood with no correlation to microalbuminuria (Caner et al., 2014). IL-33 was also recently shown to promote the recruitment of ILC2 into the white adipose tissue and promote “beiging” (Brestoff et al., 2015). Recently, IL-33 was also used to enhance Tregs and ILC2 for protection from acute renal injuries (Cao et al., 2018). However, no studies have been conducted for using cytokine based enhancement of Tregs for type 2 diabetes and/or T2DN. Although, type-2 diabetes have been considered to be associated with metabolic syndrome, the ability of the islets to lose insulin production cannot be ruled out. As observed in our study, the islets had significantly reduced expression of insulin in the islets as compared to the IL233 treated mice (Figure 2), suggesting a possible role of Tregs and suppression of inflammation in preserving islet function in type 2 diabetes, in addition to preserving renal structure and function.

Several studies have shown the benefits of Th2 response in obesity and adipogenicity (Zhang et al., 2015). IL-33 was also demonstrated in the B6.Lep^ob/ob mouse model to protect from metabolic syndrome, induce Th2 response and polarize macrophages to an M2 phenotype, but no data on Tregs was reported (Miller et al., 2010). In the same study mice lacking IL-33 receptor (ST2) had increased body weight, adipogenicity, impaired insulin secretion, and glucose regulation compared to WT controls upon feeding of high-fat diet. Foxp3⁺ Tregs have been identified in adipose tissue, where they express IL-10 and PPAR-γ (Cipolletta et al., 2012). Clinically, mixed results have been obtained for correlation of Treg-levels with obesity (Eller et al., 2011; Zeyda et al., 2011; Agabiti-Rosei et al., 2018). In another study, treatment with IL-25 induced weight loss and improved glucose tolerance, accompanied with greater accumulation ILC2 and NKT cells, eosinophils, and AAM in visceral adipose tissue (Hams et al., 2013). In the same study depletion of ILC2 in Rag1 KO mice increased weight gain and glucose intolerance, whereas adoptive transfer of ILC2 or NKT cells induced transient weight loss and stabilized glucose homeostasis in obese mice. In our study, we observed that IL233 treatment increased the accumulation of Tregs, ILC2, eosinophils and AAM in the adipose tissue of genetically obese mice. This was accompanied with a trend toward Th2 skewing and a significant increase in IL-5 production, all of which could contribute to protection from adipogenicity, restoration of glucose clearance and decreased inflammation in the adipose tissue.

It is important to note that IL233-mediated protection was far more robust when offered to younger mice, because early intervention resulted in significantly lower albuminuria compared to the late intervention (Supplementary Figure 2). If such approaches were to be adopted clinically, treatment should be targeted to early onset obesity. We also observed a robust and sustained increase of Tregs in the pancreatic lymph node as well as spleens of the mice treated IL233, which likely contributed to reduced systemic inflammation as observed with higher Treg/TNFα ratio than saline-treated controls (Supplementary Figure 4). Thus, the IL233 hybrid cytokine, which contains the activities of IL-2 and IL-33, promotes Tregs systemically, inhibits inflammation, restores glucose homeostasis, promotes recruitment of ILC2, eosinophils, and AAM in the adipose tissue to attenuate obesity, offers multi-level protection from obesity linked diabetic nephropathy. Combined with a possibility of induction of a reparative program in kidneys (Sabapathy et al., 2019), IL233 bears strong therapeutic potential.

This study was carried out in accordance with NIH Guide for the Care and Use of Laboratory Animals and approved by University of Virginia Animal Care and Use Committee.

RS conceived the project and designed the experiments. VS, MS, SM, RC, and PS performed the experiments and analyzed the data. RS, VS, and MS wrote the manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.