Itreg cells Ameliorates MOG-induced brain inflammation via endowing DC tolerogenic capacity predominantly via TGF-beta signaling mediated AKT/mTOR pathway inhibition

Abstract

Background

Certain environmental factors have been known to compromise the suppressive capacity of thymus-derived regulatory T cells (tTregs) while leaving transforming growth factor-beta (TGF-β)-induced Tregs (iTregs) unaffected. The objective of this study is to ascertain whether both Treg subsets exhibit comparable efficacy in regulating brain inflammation through the inhibition of immunogenic dendritic cells (DCs) and instead induce tolerogenic DCs.

Objectives

We aimed to delineate the different therapeutic potential roles of both Treg subsets in promoting the tolerogenic capacity of DCs and elucidate the mechanistic crosstalk between Tregs and DCs.

Methods

The clinical scores of experimental autoimmune encephalomyelitis (EAE) mice were continuously monitored, brain inflammation was assessed through hematoxylin and eosin (H&E) staining, and the presence of brain-infiltrating Th1/Th17 cells as well as splenic CD11c⁺ DCs was analyzed using flow cytometry. Additionally, a DC-T coculture assay was conducted, and the underlying mechanisms were determined by western blotting and flow cytometry.

Results

iTregs exhibit greater efficacy than tTregs in mitigating brain inflammation in both EAE and EAE provoked by a high-salt diet. iTregs suppress the pro-inflammatory activity of DCs while promoting the generation of a tolerance-inducing DC phenotype. This effect is primarily mediated by membrane-bound TGF-β signaling, rather than through IL-10R signaling, and involves the inhibition of the AKT/mTOR pathway.

Conclusion

iTreg cells play a pivotal role in orchestrating the formation of a robust immunoregulatory circuit involving tolerogenic DCs, which holds significant promise as a target for the development of innovative immunotherapeutic strategies for autoimmune disorders.

Supplementary Information

The online version contains supplementary material available at 10.1186/s10020-025-01408-x.

Introduction

CD4⁺CD25⁺Foxp3⁺ regulatory T cells (Tregs), central mediators of immune suppression, utilize cell-contact mechanisms (e.g., TIGIT, CD39/CD73) and inhibitory cytokines (IL-10, TGF-β) to modulate aberrant immunity (Raffin et al. 2020). Heterogeneous Treg subsets—thymus-derived (tTregs), peripheral (pTregs), and TGF-β-induced (iTregs)- offer distinct therapeutic potential for autoimmune diseases (Zheng et al. 2006; Chen et al. 2021). Previous studies, including our own, have underscored both the similarities and distinctions among specific Treg subsets. tTregs and iTregs exhibit distinct developmental and functional profiles. Markers like Helios (Ikzf2) and Neuropilin-1 (Nrp1) often distinguish tTregs. While Nrp1 is dispensable for general immune suppression, it is crucial for tTregs in specific contexts like limiting anti-tumor responses or resolving colitis, potentially by modulating Akt/Foxo3a signaling to enhance stability (Zhang et al. 2020; Zhou et al. 2011). A major limitation for Tregs in adoptive therapy is their instability. This stability difference is epigenetically regulated: stable Foxp3 expression in tTregs correlates with demethylation of the CNS2 enhancer region in the Foxp3 locus, facilitating transcription factor binding. Sustained IL-2 signaling can induce CNS2 demethylation and stabilize Treg subsets. Gut-derived pTregs (frequently co-expressing RORγt) also exhibit CNS2 demethylation and stability, suggesting microenvironmental factors (microbiota, metabolites, cytokines) may promote stability in vivo (Zhang et al. 2020; Zhou et al. 2011). Komatsu et al. provided evidence demonstrating that adoptively transferred tTregs preferentially lose Foxp3 expression and undergo trans differentiation into pathogenic T helper 17 (Th17) cells, thereby exacerbating the onset and severity of the CIA model (Komatsu et al. 2014). Furthermore, other scholars, in conjunction with our findings, have reported the capacity of tTregs to differentiate into other T effector cell subsets, including Th1, Th2, and Th17 cells, concomitant with the downregulation of Foxp3 expression and a concomitant reduction in their immunosuppressive functionality (Zhou et al. 2010). These changes occur in the presence of pro-inflammatory cytokines both in vitro and in vivo (Komatsu et al. 2014; Luo et al. 2019). Furthermore, our studies have documented the resistance of iTregs, but not tTregs, to IL-6-driven conversion into Th17 cells. Notably, adoptive transfer of iTregs, but not tTregs, significantly mitigated bone erosion in the CIA model, attributed to the enhanced stability and functionality of iTregs compared to tTregs following cell infusion (Kong et al. 2012; Yang et al. 2020a).

Emerging evidence has highlighted the influence of environmental factors, particularly high-salt diet, on the immunoregulatory capacity of distinct Treg subsets (Kleinewietfeld et al. 2013). Sodium chloride (NaCl) constitutes a fundamental component of daily dietary intake and plays a crucial role in maintaining bodily homeostasis. Preliminary investigations have unveiled that excessive salt promotes the differentiation of Th17 cells via the IL-23R-SGK1 signaling pathway, culminating in a highly pathogenic phenotype that exacerbates EAE (Yang et al. 2020b). High-salt intake was demonstrated that can activate damage-associated molecular patterns, the complement system, inflammasomes, leading to salt-sensitive hypertension (Maaliki et al. 2024). High concentrations of extracellular salt induce lipid oxidation and the formation of isoketal adducts in DCs, which in turn, exacerbates high blood pressure (Small et al. 2018). Dietary salt intake also induces high levels of OVA-specific serum IgG, IgG1, IgG2a, and IgE in a OVA-induced murine food allergy through promoting Th2-mediated immune responses (Liu et al. 2021). In a vascular cognitive murine model, high salt has been associated with cognitive dysfunction by attacking the cerebral microvasculature, through an adaptive response, initiated in the intestine and mediated by Th17 cells (Martín-Hersog et al. 2022). Furthermore, high salt increases the secretion of IFNγ in tTregs, compromising their suppressive functionality and exacerbating experimental graft-versus-host disease (Sumida et al. 2019). In contrast, the results of our study revealed that, unlike tTregs, iTregs remain relatively stable and function effectively under conditions of elevated sodium chloride concentrations. Notably, high salt exposure does not significantly change the transcriptional profiles of either iTreg-specific markers or pro-inflammatory genes. By using a model of high salt-induced colitis, we corroborated that iTregs exert substantial control over colitis progression, whereas tTregs predominantly lose their inhibitory potency (Luo et al. 2019).

EAE serves as the primary experimental model for multiple sclerosis (MS), a human inflammatory demyelinating disorder characterized by immune dysregulation and infiltration of immune cells into the central nervous system (CNS) (Rodríguez Murúa et al. 2022). Notably, T cells play a pivotal role in the pathogenesis of EAE, wherein peripheral T cell activation by viral or other infectious antigens or superantigens leads to the production of inflammatory cytokines and facilitates their traversal across the blood–brain barrier. The severity of EAE is also correlated with the recruitment of dendritic cells (DCs) into the CNS (Sie and Korn 2017). Particularly, plasmacytoid DCs, which are highly specialized antigen-presenting cells (APCs), assume a critical role in immune activation by bridging innate and adaptive immune systems (Piacente et al. 2022). Under immune homeostatic conditions, DCs patrol the CNS microenvironment, functioning as sentinels. Upon activation, these DCs adopt a pro-inflammatory phenotype and migrate to lymph nodes, thereby fostering the generation of self-reactive T cells and other immune cell subsets. Treg cells critically suppress immune responses by targeting DCs. Key evidence includes reduced Treg suppressive capacity when APC numbers increase in vitro, Treg-mediated downregulation of co-stimulatory molecules (CD80/CD86) on DCs/APCs, and Treg blockade of conventional T cell binding to DCs. In vivo, Treg depletion expands DC populations, indicating Treg control over DC development. Molecular mechanisms involve Treg clustering with DCs via LFA-1 and CTLA-4-mediated downregulation of CD80/CD86, achieved through trans-endocytosis/trogocytosis. This CD80 downregulation additionally enhances immunosuppression by freeing PD-L1 for interaction with PD-1. Furthermore, Treg cells deplete peptide-MHC class II complexes from DCs in a TCR-specific manner. Suppression of DC function is thus a central Treg mechanism, executed via multiple distinct pathways (Yamazaki 2024).

DCs actively mediate immune tolerance through multifaceted mechanisms, underpinning central and peripheral tolerance while resolving active immune responses. Tolerogenic DCs (tol-DCs) achieve this by restraining effector T cells and driving induced Treg cell differentiation. Critical pathways involve surface molecule engagement—notably PD-L1, which suppresses T-cell activation via Akt/mTOR inhibition to skew naive T cells toward iTreg fate, while PD-L1/PD-L2 concurrently deliver direct inhibitory signals to enforce CD4⁺/CD8⁺ T-cell tolerance. Secreted immunomodulators (IL-10, IL-27, TGF-β, retinoic acid) and the enzyme IDO further promote iTreg conversion, complemented by metabolic control through extracellular ATP/adenosine regulation (Domogalla et al. 2017; Chattopadhyay and Shevach 2013). Tol-DCs additionally enforce tolerance via direct T-cell functional modulation, including anergy induction and T-cell deletion. Pharmacologic agents (vitamin D3, corticosteroids, rapamycin) or immunoregulatory cytokines (IL-10, TGF-β) can experimentally generate tol-DCs. Notably, a bidirectional tolerogenic circuit exists: Treg cells license conventional DCs to adopt immunosuppressive functions, thereby curbing Th1/Th17 responses and mitigating autoimmune pathology (Chattopadhyay and Shevach 2013; Gao et al. 2015). Consequently, despite numerous studies documenting the preventive roles of both tTregs and iTregs in EAE progression, uncertainties persist regarding whether both Treg subsets exert equivalent immunosuppressive effects in inhibiting EAE exacerbation induced by high dietary salt levels. Additionally, elucidating whether Treg cells modulate EAE progression through DC phenotype modulation remains an important avenue of inquiry (Chattopadhyay and Shevach 2013; Griffiths et al. 2009).

While multiple studies confirm that Treg cells induce DC tolerance, it remains unclear whether different Treg subsets possess equivalent tolerogenic capacity under complex inflammatory milieus—particularly when exposed to combined stressors like high salt and IL-6. Here, we demonstrate that in vitro, iTregs exhibit striking resilience to concurrent high-salt and IL-6 exposure. Unlike tTregs, iTregs maintained tolerogenic function in this inflammatory environment, ultimately mitigating high salt-aggravated EAE progression. Our study underscores the pivotal role of iTregs in mediating the generation of tol-DCs in the context of the complex environmental factors contributing to EAE progression. The administration of iTregs holds promise as a potential therapeutic approach for patients with MS. Notably, our study represents the first instance of a comparative analysis experiment designed to concurrently assess the functional capabilities of both tTreg and iTreg subsets in modulating the biological function of DCs in the context of EAE.

Materials and methods

Mice

All the mice were purchased from The Jackson Laboratory and Lanzhou Veterinary Research Institute and were bred and housed under specific pathogen-free conditions in the animal facilities of Hershey Medical Center, Penn State University, and Lanzhou University Medical School. C57BL/6-Foxp3-GFP and B6 Rag1^−/− mice were purchased from The Jackson Laboratory, while C57BL/6 (B6) mice were obtained from Lanzhou Veterinary Research Institute. Male mice aged 6 to 10 weeks were used in all experiments.

Animal procedures and euthanasia

All animal experiments were approved by the Institutional Animal Care and Use Committee (IACUC) of (LDYYLL-2023–454) and were conducted in accordance with the AVMA Guidelines. Different anesthetic/euthanasia methods were applied based on the specific requirements of downstream assays to minimize potential confounders: For flow cytometry and biochemical analyses: Mice were euthanized by cervical dislocation under CO₂ inhalation. Spleens and CNS tissues were harvested immediately for immune cell isolation. This method was chosen to ensure rapid tissue processing and to avoid potential effects of prolonged anesthetic exposure on immune cell phenotypes and signaling pathways. For histology: Mice were anesthetized by intraperitoneal injection of 10% chloral hydrate (0.2 ml/mouse) prior to transcardial perfusion with 4% paraformaldehyde. This deep anesthetic regimen was selected to ensure a painless and stable state during the prolonged perfusion procedure. The use of chloral hydrate was specifically justified and approved by the IACUC for this terminal procedure only.

EAE, high salt diet

The anesthetized mice were subcutaneously injected at two sites on the flank with 250 μg of MOG_35–55 (Proteimax), fully emulsified in an equivalent volume of Complete Freund's Adjuvant (CFA, Sigma, St Louis, MO, USA) containing 4 mg/ml of heat-killed Mycobacterium tuberculosis H37Ra (Difco Laboratories, Detroit, MI, USA). At 0 h and 48 h post-immunization, intraperitoneal injections of 500 ng pertussis toxin (Alexis, San Diego, USA) were administered to the mice. Clinical scores were recorded daily, with the mean score of each mouse being noted every three days. In certain experiments, mice were provided with a sodium-rich chow containing 6% NaCl (ssniff, Germany; Jiangsu Xietong, Nanjing, China) and tap water containing 1% NaCl (NaCl high) ad libitum for 2 weeks prior to EAE induction. Throughout the experimental period, the weight of all animals was monitored every 3 days, with no observable differences noted (data not shown), indicating similar consumption of food and water between groups.

Adoptive transfer experiments

C57BL/6 EAE mice were randomly assigned to receive tail intravenous injections of iTregs, tTregs (both 1*10⁶ cells per mouse), or vehicle saline on day 4 post-immunization. To investigate the suppressive mechanisms of Treg subsets in vivo, anti-IL-10R (0.25 mg/kg body weight) or isotype-matched IgG1 antibody, or ALK5 inhibitor (LY-364947, 0.5 mg/mouse; Sigma) were administered intraperitoneally weekly for a total of 3 injections, with the first injection occurring the day after cell infusion. In most experiments, each group consisted of 5 mice, with experiments repeated at least twice yielding similar results, and the data presented here represent one of those experiments. On day 30 after MOG injection, mice in each group (n = 5 mice per group) were euthanized, and cells from spleens and brains were harvested. The proportions of Foxp3⁺, IL-17A⁺, or IFN-γ⁺ T cells or CD11c⁺ DCs were determined by flow cytometry.

Tissue sampling and immune cell isolation from EAE mice

At day 30, EAE mice were euthanized, and cold phosphate-buffered saline (PBS) was used for trans cardiac perfusion in anesthetized mice. Subsequently, spleens, lymph nodes, and brains were dissected for further processing. Specifically, spleens and lymph nodes were placed in cell strainers (70 μm) and homogenized using syringe plungers, followed by washing with PBS. Brain tissues were initially sectioned into pieces and then subjected to digestion using collagenase IV (Sigma Aldrich) at 37 °C with agitation for 60 min. Following digestion, the tissues were passed through cell strainers (70 μm) and subsequently separated using a 30%/70% Percoll solution (GE) via centrifugation. Flow cytometric analysis (FACS) was conducted to analyze cells at the interface.

Histology

Mice in each group underwent perfusion with 4% (w/v) paraformaldehyde under deep anesthesia induced by intraperitoneal injection of 10% chloralhydrate (0.2 ml/mouse). Subsequently, their brains were meticulously extracted and embedded in paraffin. Brain sections were dissected and subjected to staining with hematoxylin and eosin (H&E) following established protocols. Animals were killed using the perfusion with 4% (w/v) paraformaldehyde under deep anesthesia (10% chloralhydrate, 0.2ml/mouse, i.p.). Brains and spinal cords were removed, immersed in 10% formalin, and then embedded in paraffin. Sections were then dissected and stained with hematoxylin & eosin (H&E) to assess degree of inflammatory cell infiltration according to standard protocols. Overall brain inflammation is scored from 0 to 4 based on the most severe pathology observed: 0, indicates no inflammatory cells in subpial or perivascular areas; 1, represents mild pathology, defined as minor subpial infiltration or 1–2 perivascular cuffed vessels per section; 2, denotes moderate pathology, characterized by numerous subpial infiltrating cells or 3–4 perivascular cuffed vessels per section; 3, reflects severe pathology, involving majority pial involvement (subpial) or 5–8 perivascular cuffed vessels per section; 4, signifies maximal pathology, defined as entire pial involvement (subpial) or > 8 perivascular cuffed vessels per section (Griffiths et al. 2009).

tTreg isolation, iTreg generation, Treg subset conversionin vitro and in vivo

tTregs were isolated from the thymus of WT B6 mice or Foxp3^GFP B6 mice through gating of CD4 + CD25^high or CD4⁺GFP⁺ cells, achieving a purity of 99%. These cells were subsequently stimulated with anti-CD3/CD28-coated beads (Invitrogen, Carlsbad, CA) at a ratio of 1 bead:2 cells and rhIL-2 (100 U/ml; R&D Systems, Minneapolis, MN) for 72 h. For iTreg generation in vitro, naïve CD4⁺CD44⁻CD62L⁺ T cells from WT or Foxp3^−GFP B6 mice (purity > 95%, CD4-Percp-Cy5.5, CD25-PE, and CD62L-APC) were stimulated with anti-CD3/CD28 dynabeads (cells: beads = 5:1, Invitrogen) in the presence of 50 U/ml rh-IL2 and 2 ng/ml rhTGF-β (both from R&D System) for 72 h. Both Treg subsets were cultured in 96-well U-bound plates in X-VIVO 15 medium (LONZA) in the presence or absence of NaCl (40 mM, Sigma-Aldrich), IL-6 (100 ng/ml), or a combination of both (Kong et al. 2012).

For in vivo Treg subset conversion, tTregs were isolated from the thymus, while iTregs were generated from splenic naïve CD4⁺T cells, both sourced from B6 Foxp3-GFP mice. Subsequently, to ensure purity, cells were sorted into CD4⁺GFP⁺ cells with a purity of 99%. A total of 0.5 × 10⁶ cells were intravenously injected into Rag1^−/− mice (C57BL/6). In some groups, tTregs or iTregs were pre-treated with NaCl (40 mM) and IL-6 (100 ng/ml, R&D System) for 48 h prior to sorting and adoptive transfer. Mice were euthanized on days 5 and 10, and T cells from the mesenteric lymph nodes (MLN) were stained for Foxp3-GFP, IFN-γ, and IL-17A.

Treg suppression assayin vitro

Initially, both Treg subsets were pre-incubated with or without NaCl (40 mM) and IL-6 (100 ng/ml) in the presence of anti-CD3/CD28 microbeads (5 cells per bead) for 3 days. Subsequently, the cells were co-cultured with naïve T cells (5 × 10⁵) stained with 2 μM carboxyfluorescein succinimidyl ester (CFSE) at various ratios as indicated. The cell mixtures were then stimulated with soluble anti-CD3 (0.5 μg/ml) for 3 days. The proliferative levels of CFSE-CD4 + T cells were assessed by assessing the rates and intensity of CFSE dilution using flow cytometry.

Co-culture of naive T responder cells with DC subsets

Splenic CD11c⁺ DCs were isolated from PBS-, iTreg-, tTreg-treated EAE mice at day 30 using the mouse CD11c MicroBeads isolation kit (Miltenyi Biotec); 0.1*10⁵ PBS-DCs, iTreg-DCs or tTreg-DCs were co-cultured with 0.5*105 naive T responder cells in 96-well U-bottom plates.

For DC-mediated proliferation assay, the WT B6 naïve T cells (CD4⁺CD62L⁺CD44⁻) were labeled with CFSE (2 uM, BioLegend), and the co-culture system was stimulated with anti-CD3 (0.1 ug/ml; BD Pharmingen). T-cell proliferation was assessed by analyzing the CFSE dilution rates after 3 days of culture.

For the DC-induced Treg cell differentiation assay, naïve T cells (CD4⁺CD62L⁺CD25⁻CD44⁻) were sorted from the B6 Foxp3-GFP mice. These cells were then stimulated with anti-CD3 (0.5 ug/ml) and rh-IL2 (50 U/ml) for 3 days. The GFP expression was assessed using flow cytometry.

Co-culture of Treg cells with splenic DCs

Magnetic-activated cell sorting (MACS) was employed to isolate CD11c⁺ cells from the spleen, which were subsequently cultured in 24-well plates using a medium supplemented with GM-CSF (500 U/ml) and IL-4 (200 U/ml). After 3 days of culture, cells were collected, washed, and re-seeded in 48-well plates containing RPMI 1640 medium supplemented with 10% fetal calf serum and 1% essential amino acids, along with anti-CD3 (0.5 μg/ml; BD Pharmingen). Subsequently, tTregs or iTregs were added to DCs at a 2:1 T to DC ratio. Following 12 h of coculture, T cells were depleted through MACS separation. In some wells, tTregs and iTregs were pre-treated with 40 mM NaCl for 48 h, then washed and added to the wells.

Flow-cytometric analyses

Cells from the spleens and the brains were stained with monoclonal antibodies (mAbs) and isotype controls for CD3(OKT3, 12–0037–42), CD4(RM4-4, 12–0043–82), and CD25(PC61.5, 17–0251-82) (all from eBioscience). For intracellular staining of cytokines, cultured cells were stimulated with Phorbol 12-Myristate 13-Acetate (PMA) and ionomycin (both at 0.25ug/ml; Sigma-Aldrich) for five hours at 37 °C in the presence of brefeldin A (5 ug/ml; BioLegend) for the last 4 h. After surface staining, cells were fixed and permeabilized using the Transcription Factor Staining Buffer Set (Thermo Fisher Scientific) and stained intracellularly using the following antibodies: anti-IL-17-APC (TC11-18H10, 130–103-007) and anti-IFNγ-PE (AN.18.17.24, 130–123–700).

For DC staining, total splenocytes were treated with anti-CD16/CD32 antibodies (Ab93) (BD Biosciences, 567020) and then stained with the following surface antibodies: anti-CD11c (N418, 12–0114-83), anti-MHCII (M5/114.15.2, A14902), anti-CD80 (16-10A1, 17–0801-82), anti-CD86 (GL1, 17–0862-82), anti-LAP (TW7-16B4, 25–9821-82), anti-B7H4 (188, 12–5972-82), anti-IDO (mIDO-48, 50–9473-82) (all from eBioscience), and anti-IL-10 (JES5-16E3) (Biolegend, 505003). IL-10 production was measured by stimulating total splenocytes with 10 ng/ml PMA and 100 µg/ml ionomycin for 6 h in the presence of 5 µg/ml Monensin. After surface CD11c staining, cells were fixed, permeabilized, and then anti-mouse IL-10 was used for intracellular staining. All flow cytometric analyses were conducted using the following isotype controls: Hamster IgG1-PE (G235-2356), rat IgG2a-APC (R35-95). Data were evaluated by FACS Aria II (BD Biosciences, San Jose, CA) and analyzed by FlowJo software (Tree Star).

Western blotting

Total protein was extracted from cultured CD11c⁺ cells and transferred onto PVDF membranes, which were subsequently blocked in 5% BSA or 5% non-fat milk in TBST for 1 h. The membranes were then incubated overnight at 4 °C with antibodies against phospho-Akt (Ser473), total AKT, phospho-P70S6 kinase (Thr389), total P70S6 kinase, and GAPDH. All antibodies were procured from Cell Signaling Technology (Danvers, MA). On the following day, the membranes were incubated with HRP-conjugated secondary antibodies for 1 h. Protein bands were detected using Enhanced Chemiluminescence (ECL) Reagent (BeyoECL Star, Beyotime Biotechnology, Shanghai, China) and visualized with a chemiluminescence imaging system (e.g., Tanon 5200 Multi or Bio-Rad ChemiDoc™ MP).

Statistical analysis

Statistical analysis was conducted using Graph Pad Prism Version 5 software and presented, if not indicated elsewhere, as mean ± SEM. To assess differences between groups, Student’s t-test was used. For comparisons involving three or more groups, one-way ANOVA analysis was performed. A p < 0.05 was considered statistically significant (*p < 0.05; **p < 0.01; ***p < 0.001; ns, not significant).

Results

iTregs are more stable and have superior functional qualities compared to tTregs in response to various stimulants

Previously, it has been reported that the inflammatory cytokine IL-6 drives the conversion of tTregs, rather than iTregs, into pathogenic Th17 cells (Sakaguchi et al. 1995; Zheng et al. 2002). Additionally, sodium chloride converts tTregs, but not iTregs, into IFNγ-producing Th1-like cells. To further compare the conversion of the two Treg subsets into the Th17 or/and Th1 lineages, NaCl combined with IL-6 was used as a stimulant. tTregs (CD4⁺GFP⁺ cells) were sorted from thymic cells in Foxp3-GFP mice, while iTregs were induced from splenic naïve CD4⁺CD44⁻CD62L⁺GFP⁻ cells using a standard protocol as previously described (Figure S2A) (Luo et al. 2019). Initially, these Tregs were cultured in the presence of NaCl (40 mM) or IL-6 (100 ng/mL) for 3 days, and subsequently, IFNγ- or IL-17A-producing cells were examined by FACS. The gating strategies are provided in Supplementary Fig. 1. Reflecting the established notion that iTregs possess superior stability relative to tTregs, our analysis revealed that tTregs exhibited a significant increase in the frequency of IL-17A- producing cells upon exposure to either a NaCl or IL-6 compared to fresh conditions, with the increases being statistically significant (P < 0.05 and P < 0.001, respectively), whereas the frequency of IFNγ- producing tTregs, while showing an upward trend, did not differ significantly across the three conditions (Figure S2B-D). By contrast, the frequencies of both IFNγ- and IL-17A- producing iTregs remained unchanged and showed no significant differences among the groups, collectively demonstrating the functional instability of tTregs and the remarkable stability of iTregs under these challenging conditions (Figure S2B-D).

Subsequently, we assessed the stability of these two Treg subsets in vivo. Both tTregs and iTregs were pretreated with IL-6 and NaCl for 48 h and were then adoptively transferred into Rag1^−/− recipients at a dose of 0.5*10⁶ per mouse to assess their stability under lymphopenic conditions for 10 consecutive days. GFP-expressing cells were quantified in the MLN on day 5 and day 10, respectively. As depicted in Fig. 1C and D, there was some loss of Foxp3 in both Treg populations, but slightly less in iTregs than in tTregs at day 5 after adoptive infusion. However, by day 10, approximately 75% of Foxp3 expression was retained in iTregs, especially in those pretreated with NaCl and IL-6. Conversely, the pretreated tTregs displayed a notable reduction in Foxp3 expression compared to untreated tTregs. Concurrently with changes in Foxp3 expression, both tTregs and pretreated tTregs exhibited robust expression of IL-17A and IFNγ, with the production of these two cytokines significantly higher in pretreated tTregs than in untreated tTregs (Fig. 1E, F). In line with the in vitro findings, no significant expression of IL-17A and IFNγ was observed in either iTregs or pretreated iTregs (Fig. 1E, F).

Fig. 1.

High salt levels have different effects on Foxp3 expression and conversion of both Treg subsets in vitro and in vivo. tTregs were isolated from the thymus of C57BL/6 Foxp3-GFP knock-in mice and subsequently expanded and activated for 72 h. Concurrently, iTregs were generated from naive CD4⁺ T cells in the presence of anti-CD3/28 microbeads, IL-2, and TGFβ in standard media for 72 h. Following this, both Treg subsets were harvested, thoroughly washed, and then cultured in media supplemented with NaCl (40 mM) and IL-6 (100 ng/mL) for an additional 48 h, denoted as pretreated-tTregs or pretreated-iTregs. Subsequently, these cells were collected, surface-stained for CD4, and intracellularly stained for IL-17A and IFN-γ, with measurements performed on CD4⁺GFP⁻ and CD4⁺GFP⁺ cells via FACS. Representative FACS plots from three independent experiments are depicted (A, B), (n = 6). In a parallel experiment, tTregs, pretreated-tTregs, iTregs, and pretreated-iTregs were transferred into Rag1^−/− mice. The mice were euthanized at day 5 or day 10, and total MLN cells were collected. Assessment of Foxp3-GFP loss (C, D) and determination of the frequency of CD4⁺ IL-17A⁺ and CD4⁺ IFN-γ⁺ cells (E, F) in the MLN from each group were compared. Representative results, presented as mean ± SEM from two distinct experiments. Furthermore, iTregs were restimulated under 40 mM NaCl for 3 days and subsequently analyzed by flow cytometry for IL-17A and IFN-γ. Treg subsets were co-cultured with responder CD4.⁺ T cells at variuos ratios; representative FACS data (G) and quantification (H) are shown. The bar graph summarizes results from independent experiments (n = 3). Statistical analysis was conducted using paired t-test (A, B), one-way ANOVA (C- F), two-way ANOVA (G, H). Notably, ns denotes not significant, * indicates p < 0.05, ** indicates p < 0.01, *** indicates p < 0.001, **** indicates p < 0.0001

Finally, we sought to ascertain whether the two Treg subsets retained their function faithfully when pretreated with IL-6 and NaCl. To assess functionality, Treg subsets primed with/without (NaCl + IL-6 control) for 3 days were cultured with CFSE-labeled splenic CD4⁺ T cells (T responder cells) at different ratios (1:1, 1:4). Upon addition of iTregs or tTregs to T responder cells, CFSE labeling revealed a marked inhibition of proliferation, demonstrating a dose-dependent inhibitory effect. Nevertheless, tTregs subjected to pretreatment displayed a marked reduction in their inhibitory function, whereas iTregs that were pretreated preserved their suppressive capacity (Fig. 1G, H). In summary, these results unequivocally demonstrate that iTreg cells are more stable and exhibit superior functionality compared to tTregs under conditions of inflammatory cytokines combined with sodium chloride.

iTregs exhibit superior functional activity compared to tTregs in EAE-mediated brain inflammation

To investigate the distinct effects of the two Treg subsets in mitigating brain inflammation, we used the MOG_35–55-induced EAE model. Disease manifestation in mice was observed at day 12 post-immunization, characterized initially by tail weakness, followed by hindlimb and forelimb paralysis, as previously described (Zheng et al. 2007; Moreau et al. 2022). Intravenous transfer of Treg subsets 4 days after the induction of EAE. As anticipated, both iTregs and tTregs exhibited a significant inhibitory function on the clinical expression of EAE, delaying onset and ameliorating symptoms (Fig. 2A). Brain H&E staining revealed a markedly lower level of lymphocyte infiltration in the tTreg- or iTreg-treated groups compared to the disease model group, with iTreg groups revealing a more pronounced alleviation of lymphocyte infiltration compared to tTreg groups (Fig. 2B). Notably, we observed that iTregs displayed a superior effect on restraining lymphocytic infiltration compared to tTregs, correlating with the clinical symptoms depicted in Fig. 2A. Furthermore, we analyzed the change of Th1 and Th17 cells in each group, as these subsets represent crucial pathogenic immune cells in EAE development. Flow cytometric analyses demonstrated significant suppression of Th17 cells in the spleens (Fig. 2E) and brains (Fig. 2G) of both treatment groups compared to disease model controls. Particularly, the Th17 population in brains was dramatically decreased in mice administered with iTregs compared to those administered with tTregs (p < 0.05) (Fig. 2E, G), indicating a distinct therapeutic difference between iTregs and tTregs. Subsequently, the frequency of Th1 cells in the spleens and brains was significantly reduced in iTreg-treated, but not in tTreg-treated group compared to the model group (Fig. 2F, H). Specifically, tTreg administration only inhibited brain Th1 cell development, with less obvious effects compared to iTreg administration (Fig. 2H).

Fig. 2.

iTregs outperform tTregs in terms of functional activity during EAE-induced brain inflammation. 1 × 10⁶ tTregs or iTregs, generated as described previously, were adoptively transferred into B6 male mice on day 4 following MOG_35–55/CFA immunization. Mean clinical scores on days 0–30 of each experimental group (n = 5 per group) (Panel A). Histological analyses were conducted, and the sections of H&E staining of brain tissues from each experimental group (20x) (Panel B). Experiments were terminated on day 30, following which cells from spleens (Panel C) and brains (Panel D) were harvested and stimulated with PMA, ionomycin, and BFA. Subsequently, IFNγ- and IL-17A-producing CD4⁺ T cells from spleens (Panels E, F) and brains (Panels G, H) were analyzed via flow cytometry, the plot data were gating on CD4. Data presented are representative of at least 5 mice per group. Additionally, staining of indicated CD11c⁺ splenic DCs on day 30 was performed via flow cytometry (Panels I, J). All data are representative of at least 3 independent experiments. Statistical analysis for clinical scores (A) was performed using repeated-measures ANOVA; All other data were conducted using one-way ANOVA, with significance denoted as * for p < 0.05, ** for p < 0.01, *** for p < 0.001, and ns indicating not significant

Given the critical role of DC priming of T lymphocytes in EAE pathogenesis, we analyzed splenic CD11c⁺ DCs in each group. Both tTreg and iTreg subsets decreased the frequency of splenic CD11c⁺ DCs, with iTregs displaying superior efficacy compared to tTregs (Fig. 2I, J). These findings collectively support the notion that both tTregs and iTregs alleviate EAE severity, with iTregs exhibiting more potent suppressive activity, potentially involving DCs in this effect.

iTregs inhibit the pro-inflammatory DC phenotype and promote tolerogenic activities in the EAE environment

DC activation reflects a balance of pro- and anti-inflammatory signals in EAE. Pro-inflammatory DCs express cytokines and chemokines that contribute to the pathogenesis of EAE, while tolerogenic DCs play a modulatory role in restoring immune homeostasis. Subsequently, we assessed the immunophenotype and function of splenic CD11c⁺ DCs from the three groups: PBS-treated (model-DCs, m-DCs), tTreg-treated (t-DCs), and iTreg-treated (i-DCs) EAE mice, using flow cytometry. The gating strategy was shown in supplementary Fig. 3. The median fluorescence intensity (MFI) of the co-stimulatory molecule CD80 was significantly lower only in the i-DCs compared to the m-DCs and t-DCs (Fig. 3A, B). Likewise, the MFI of another co-stimulatory molecule CD86 in CD11C⁺ DCs was significantly decreased in mice treated with the iTreg subset versus the control and tTreg-treated mice. Interestingly, i-DCs expressed much lower levels of CD86 compared to t-DCs (Fig. 3A, C). Investigation of well-known tolerogenic DC cells also revealed that i-DCs significantly expressed higher levels of latency-associated peptide (LAP) (Fig. 3A, D) and intracellular IL-10 (Fig. 3E) compared to that of m-DCs, as indicated by MFI or frequency respectively (p < 0.05). We did not find significant expression of the two tolerogenic markers in t-DCs, but only observed an increasing trend (Fig. 3A, D, E). These findings indicate that iTregs are more capable of inducing DC immune tolerance than tTregs, which may help to explain why the iTreg subset is more effective than tTregs in treating EAE.

Fig. 3.

In the EAE environment, iTregs suppress the pro-inflammatory DC phenotype while inducing tolerogenic activities. PBS-treated, tTreg-treated, and iTreg-treated EAE mice (n = 5 per group) were euthanized on day 30. Total splenocytes were collected, and DCs were stained and gated on CD11c. Expression levels of CD80, CD86, and LAP on CD11c splenic DCs were depicted (Panel A), with corresponding statistical graphs presented (Panels B, C, D). The data revealed are representative of at least 3 independent experiments. Some splenocytes were stimulated with PMA, ionomycin, and monensin, followed by staining for CD11c and IL-10. The bar graph illustrates the frequency of CD11c⁺IL-10⁺ DCs (Panel E). Additionally, PBS-treated DCs, tTreg-treated DCs, and iTreg-treated DCs were sorted from the spleens of EAE mice, then added to naive B6 T cells (labeled with CFSE) at a 5:1 T cell:DC ratio. T cell proliferation was assessed by CFSE dilution after 3 day coculture (Panel F), with statistical data indicated (Panel G). Furthermore, a portion of these three groups of DCs was added to naive T cells (from B6 Foxp3-GFPmice) at a 1:5 ratio in the presence of anti-CD3 and rhIL-2 for 3 days, and Foxp3-GFP expression was measured using flow cytometry (Panels H, I). All data presented are representative of at least 3 independent experiments. Statistical analyses were conducted using one-way ANOVA, with significance denoted as * for p < 0.05, ** for p < 0.01, *** for p < 0.001, and ns indicating not significant

Based on these findings, we hypothesize that i-DCs would be able to suppress T cell proliferation and promote Treg cell differentiation. Splenic DCs were sorted from EAE mice. As expected, when the i-DC subset was co-cultured with naive T cells (CD4⁺CD62L⁺CD44⁻) labeled with CFSE, the proliferation of CFSE⁺ T cells was significantly restrained compared to that co-cultured with m-DCs or t-DCs (Fig. 3F, G). Similarly, i-DCs also exhibited a remarkable ability to induce the development of Treg cells. Approximately 10% of the naïve T cells became CD25⁺GFP⁺ regulatory T cells (Fig. 3H, I). This was not observed in the PBS-DCs co-culture system. Although t-DCs also obviously generated a significant amount of Foxp3⁺ cells, i-DCs had a robustly powerful ability in the induction of naive T cells into Treg cells compared to t-DCs (Fig. 3I). Thus, these results indicate that iTreg-modified DCs are rendered hyporesponsive and have anti-inflammatory function. The potential therapeutic effect of the iTreg subset in the immune-mediated EAE model is likely linked to their effects on the function of DCs in vivo.

iTregs endow DCs with a tolerogenic phenotype andfunction mainly via TGF-β but not IL-10 signaling

It has been well documented that both Treg subsets secrete soluble TGF-β and IL-10, which are crucial for exerting immunosuppressive function. To further determine whether membrane-bound TGF-β and IL-10 receptors are also required for Treg-mediated suppression in EAE, we used an inhibitor of TGF-β receptor I (ALK5) and an anti-IL-10R antibody in the EAE model. EAE induction and iTreg cell infusion were performed, as depicted in Fig. 2A. Some of the mice treated with iTreg cells were injected with a neutralizing anti-IL-10R antibody (isotype-matched IgG1 antibody as its control) or ALK5 inhibitor (ALK5i) ad depicted in supplementary Fig. 4. Our results revealed that iTreg infusion exerted a significant inhibitory function on the treatment of EAE, as demonstrated by a reduction in clinical scores (Fig. 4A) and histological brain inflammation (Fig. 4B). Additionally, administration of DMSO or IgG1 alone did not influence disease progression in the EAE mice compared to the model group (Figure S5). The blockade of IL-10 signaling via the anti-IL-10R antibody resulted in a statistically significant effect on the therapeutic efficacy of iTregs, as determined by clinical scores. Nevertheless, the unchanged neuroinflammation scores following anti-IL-10R blockade suggest that IL-10 signaling may play a limited role in the immunosuppressive capability of iTreg cells in the treatment of EAE (Fig. 4B). Conversely, the clinical scores and the pathological inflammation scores were both completely reversed in the iTreg + ALK5i group compared to the iTreg only group, indicating that blockade of TGF-β receptor I activity apparently impaired the protective effect of iTreg cells (Fig. 4A, B). These results preliminary indicate that TGF-β and IL-10 signaling both play a role in executing the immunosuppressive function of iTreg cells, whereas TGF-β signaling is more crucial. The presence of Th17 and Th1 cells in the brains was also examined using flow cytometry. As expected, iTreg infusion significantly repressed the brain Th17 cell differentiation. However, co-administration of ALK5i mostly restored the Th17 frequency in the brains to EAE levels compared to the iTreg infusion group. At the same time, we only observed the reverse trend in the inhibition of Th17 cell frequency by iTregs treatment in the anti-IL-10R antibody administration group (Fig. 4C, D). Also, IFN-γ-producing Th1 cells were also statistically decreased in iTreg-treated brains. Conversely, neither ALK5i nor anti-IL-10R could significantly reverse the inhibitory effect on Th1 differentiation by iTreg infusion (Fig. 4C, E). These observations indicate that TGF-β signaling plays a dominant role in the immunosuppressive function of iTregs on EAE, whereas IL-10R signaling plays a partial role, if any.

Fig. 4.

iTregs give DCs a tolerogenic phenotype and function primarily through TGF-β, not IL-10 signaling. iTreg cells were generated and administered, and EAE was induced as described previously. Subsequently, iTregs, ALK5i, or anti-IL-10R were administered in separate groups, with DMSO treated as the model group. The severity of EAE was assessed. Mean clinical scores on days 0–30 of each experimental group (n = 5 per group) (Panel A). Histological changes in the brains of mice from each group on day 30 post-immunization were assessed by removing brain sections, fixing them, and conducting H&E staining, with typical photographs displayed (Panel B). On day 30 post-immunization, brains were harvested, and single-cell suspensions were prepared. Populations of IL-17-producing Th17 cells (Panels C, D) and IFN-γ-producing Th1 cells (Panels C, E) were analyzed by flow cytometry, with representative flow cytometry data indicated for each group. Additionally, the frequency of splenic CD11c⁺ DCs was examined by flow cytometry (Panel F, G). Splenic CD11c⁺ DCs were sorted from each group on day 30, and some were co-cultured with naive Foxp3-GFP CD4.⁺T cells (Panel H), Foxp3-GFP expression (Panel I) were assessed by flow cytometry after 3 days of culture. In vitro results presented as means ± SEM of triplicate wells from three independent experiments, with n = 5 mice per group. Statistical analyses: repeated-One-way ANOVA (A), One-way ANOVA (B-I); Notably, ns denotes not significant, with * indicating p < 0.05, ** indicating p < 0.01, *** indicating p < 0.001, and ns indicating not significant (PBS-DC vs. iTreg-DC), (iTreg-DC vs. (iTreg + ALK5i)-DC), (iTreg-DC vs. (iTreg + anti-IL-10R)-DC)

In line with the aforementioned findings, we examined the frequency of splenic DCs from each group of EAE mice at day 30 post-immunization. iTreg-treated mice revealed a remarkable decrease in CD11c⁺ DCs in spleens. ALK5i administration completely diminished this phenomenon, but anti-IL-10R did not (Fig. 4F, G). CD11c⁺ DCs expressed a baseline level of CD80, CD86, and LAP in splenocytes treated with PBS. DCs from iTreg-treated mice exhibited a downregulation of CD80 and CD86 (Figure S6A, S6B), along with an upregulation of LAP expression (Figure S6A, S6C). In comparison to the iTreg-treated splenocytes, co-administration of ALK5i significantly upregulated the expression of CD80 and CD86 (Figure S6A, S6B), while decreasing the expression of LAP (Figure S6A, S6C). However, we did not observe any statistically significant changes in these molecules in the co-administration of the anti-IL-10R group (Figure S6A-6C). Next, a series of DC-T cell co-culture experiments were conducted to assess the impact of ALK5i and anti-IL10R on the tolerogenic activity of these in vivo-modified DCs. CD11c⁺ DCs were sorted from the splenocytes of each EAE group on day 30. These DCs were cocultured with B6 Foxp3-GFP naïve T cells as described in Fig. 3. As revealed in Fig. 4H, compared to DCs sorted from PBS-treated EAE mice, iTreg-DCs (i-DC) enabled up to 11.9% of naive T cells to begin expression of Foxp3. However, splenic DCs from the ALK5i co-administration group dramatically lost their tolerogenic activity, resulting in little formation of CD4⁺Foxp3⁺ cells. However, DCs from the anti-IL10-treated group exhibited some degree of tolerogenic function, although the generation of CD4⁺Foxp3⁺ cells were statistically lower compared to iTreg-DCs (Fig. 4H, I). Taken together, these data reveal that iTregs primarily induce the formation of tolerogenic DCs via membrane-bound TGF-β signaling, rather than IL-10 signaling.

Inhibition of the AKT/mTOR pathway causes iTreg cells to gain DC tolerogenic capacity.

The involvement of the AKT/mTOR signaling pathway in the inflammatory processes of DCs has been exhaustively demonstrated. Phosphorylation of AKT leads to the activation of mTOR and its downstream targets. P70S6K, one of the most important substrates of mTOR signaling, contributes to the regulation of inflammatory cytokine production in DCs, participating in their antigen presenting and proinflammatory function (Gao et al. 2015; Chen et al. 2020). To investigate whether iTregs modify the immunosuppressive capacity of splenic DCs via AKT/mTOR signaling, we compared their intracellular levels, including total AKT (t-AKT), phospho-AKT (p-AKT), total P70S6K (t-P70S6K), and phospho-P70S6K (p-P70S6K) in splenic DCs treated with iTregs or tTregs. As measured by western blot, the expression of total AKT and P70S6K was similar in DCs treated with iTregs or tTregs (Fig. 5A, B). However, p-AKT and p-P70S6K decreased considerably only in the iTreg-treated group, and these differences were statistically significant (Fig. 5A, C). Given that high salt and/or inflammatory cytokines distinctly influence the stability and function of both Treg subsets in vitro and in vivo, we immediately assessed the phenotype of DCs cocultured with the Treg subsets that had been pretreated with a combination of NaCl and IL-6. Pretreated iTregs significantly inhibited p-Akt and p-P70S6K expression on DCs compared to that of pretreated tTregs (Fig. 5A, E). We did not find statistical differences in t-Akt and t-P70S6K between the two groups (Fig. 5A, D). These findings suggests that the regulation of the AKT/mTOR signaling pathway plays a pivotal role in mediating the suppressive effects of iTregs treatment on the immunomodulatory activity of DCs. Additionally, it can be elucidated that the iTreg subset exhibits enhanced suppressive activity in facilitating the tolerogenic function of DCs in comparison to the tTreg subset. This superior suppressive capability may be achieved through the inhibition of the AKT/mTOR signaling pathway.

Fig. 5.

Inhibition of the AKT/mTOR pathway causes iTreg cells to gain DC tolerogenic capacity. Both subsets of Tregs were stimulated with or without NaCl at a concentration of 40 mM, in the presence of anti-CD3/CD28 microbeads (5 cells per bead) and rhIL-2 at a concentration of 50 U/ml, for a duration of 3 days. Subsequently, the cells were washed and cocultured with splenic CD11c⁺ DCs for an additional 48 h, at a T to DC ratio of 2:1 (Panel A). Representative immunoblots depicting the expression levels of Akt, phosphorylated Akt (p-Akt), P70S6K, and phosphorylated P70S6K proteins are displayed (Panels B-E). The data presented are representative of three independent experiments, with statistical significance determined using a paired t-test. Significance levels are denoted as * for p < 0.05, ** for p < 0.01, *** for p < 0.001, and ns indicating not significant

iTregs, but not tTregs, have a therapeutic effect on EAE mice by reducing high-salt diet-induced brain inflammation

Previous studies have identified a population of pro-inflammatory tTreg cells that are modified by high salt. These cells are characterized by the secretion of IFN-γ and have been found to be dysfunctional both in vitro and in vivo. However, we demonstrated that TGF-β-induced iTreg populations are highly stable and functional under high salt conditions (Luo et al. 2019). Therefore, to determine the varying immunosuppressive capabilities of the two Treg subsets in regulating high-salt diet-fed EAE mice, we conducted an adoptive transfer study using a therapeutic regimen. As depicted in Fig. 6A, we observed that iTregs exerted a significant inhibitory function on the treatment of high salt fed EAE, leading to an amelioration of the clinical scores. However, tTregs had a significantly reduced inhibitory effect on high salt-fed EAE mice progression (Fig. 6A). Similarly, results from intracellular staining of IL-17A⁺ and IFNγ + of CD4⁺ T cells in the brains revealed that iTregs significantly reduced the frequency of Th1 and Th17 cells, whereas tTregs failed to suppress both inflammatory T effector cells (Fig. 6B, C). Next, as a matter of course, we examined the frequency and phenotype of splenic DCs from each group. CD11c⁺ DCs significantly decreased in the group that received iTreg administration compared to the PBS controls. However, the administration of tTreg had only a slight impact on the frequency of splenic DCs (Fig. 6D, E). Meanwhile, iTregs induced a significant decrease in the expression of CD80 and CD86 on splenic DCs. In contrast, the MFI index of these two DC maturation markers returned to levels similar to those in the PBS control group after treatment with tTregs (Fig. 6F, G). Furthermore, LAP expression was augmented in splenic DCs in the iTreg-treated group but not in the tTreg-treated group (Fig. 6F, G). Taken together, based on a face-to-face comparison experiment, we validated that both Treg subsets exhibit distinct biological characteristics in a complex environment, such as a high salt environment and under inflammatory conditions. Thus, TGF-β-induced iTreg cells may have some advantages in treating autoimmune and inflammatory diseases.

Fig. 6.

iTregs, but not tTregs, have a therapeutic effect on EAE mice by reducing high-salt diet-induced brain inflammation. Wild-type mice were subjected to a high-salt diet for two weeks prior to active immunization with MOG_35–55 peptide. Subsequently, both subsets of Tregs were generated and administered as described previously. The mean clinical scores of EAE from each group are depicted (Panel A). CD4⁺ T cells from the brains were analyzed on day 30, with flow cytometric analysis conducted to determine the frequencies of IL17A⁺ and IFNγ⁺ CD4 + cells in the respective mouse groups (Panel B, C). The frequency of splenic CD11c⁺ DCs was assessed by FACS on day 30 (Panel D, E). Kinetic analysis of CD80, CD86, and LAP expression in splenic CD11c⁺ cells was performed using flow cytometry (n = 5 per group and time point) (Panel F, G). Statistical analyses were conducted using one-way ANOVA, with significance indicated as * for p < 0.05, ** for p < 0.01, *** for p < 0.001, and ns indicating not significant

Discussion

While the initial identification of Tregs occurred within the thymus, (Sakaguchi et al. 1995) subsequent research rapidly revealed that this cell population can be induced and differentiated from non-Treg cells in the presence of cytokines such as TGF-β and interleukin-2 (IL-2) (Zheng et al. 2002, 2007). Although consensus regarding the defining characteristics that differentiate tTregs from iTregs remains elusive, both subsets have emerged as promising therapeutic modalities for autoimmune diseases and transplantation. TGFβ initiates the phosphorylation and activation of Smad2 and Smad3, pivotal for the induction of Foxp3 during iTreg generation (Moreau et al. 2022). Both Treg subsets express canonical Treg markers, such as Foxp3, CD25, GITR, and CTLA4. However, tTregs exhibit higher expression levels of PD-1, neuropilin 1 (Nrp1), Helios (Ikzf2), and CD73 compared to their iTreg counterparts (Lin et al. 2013; Yang et al. 2019). Certain studies have proposed that Foxp3 expression stability in tTregs is maintained via demethylation of CpG islands within the conserved non-coding sequence 2 (CNS2) region of the Foxp3 locus. This region serves as a binding site for various transcription factors such as Stat5 and Runx1/Cbfb, contributing to tTreg stability. Conversely, unstable Foxp3 expression in iTregs is purportedly associated with pronounced demethylation of CNS2 (Du et al. 2021). However, observations from our study and those of others challenge the notion of tTreg stability, particularly under arthritic and inflammatory conditions. Under such circumstances, tTregs exhibit susceptibility to redifferentiation into alternative T effector cell subsets, accompanied by functional changes. In contrast, iTregs demonstrate a lack of this plasticity and display enhanced suppression of osteoclastogenesis and bone erosion compared to tTregs (Zhang et al. 2020; Lu et al. 2014).

High salt levels can promote the differentiation of pathogenic Th17 cells while concurrently dampening the suppressive capacity of tTregs, thereby expediting the onset of EAE (Luo et al. 2019; Kleinewietfeld et al. 2013). By using a high-salt diet in a Rag1-/- colitis model, we provided supplementary evidence indicating that iTregs, but not tTregs, significantly ameliorate intestinal inflammation (Luo et al. 2019). Additionally, we present compelling evidence demonstrating the greater stability of iTregs compared to tTregs, as evidenced by the heightened resistance of iTreg destabilization in the presence of exogenous IL-6 and NaCl combination stimulation in vitro. In the specific EAE model used in our study, tTregs effectively curtailed disease progression. However, their efficacy was compromised in inhibiting EAE advancement in mice subjected to a high-salt diet. Notably, iTregs exhibited superior efficacy in mitigating disease progression, even under high-salt or pro-inflammatory conditions.

The therapeutic efficacy of iTregs can be elucidated by several factors. Firstly, the suppressive function of tTreg cells may be compromised by pro-inflammatory cytokines owing to their expression of IL-6R (Zhou et al. 2010). Numerous studies have reported the abrogation of tTreg suppressive activity in response to IL-6 (Khantakova et al. 2022; Gao et al. 2015). Notably, iTregs demonstrate diminished expression of IL-6 receptor compared to tTregs, rendering them resistant to IL-6-induced stimulation and preserving their phenotype and function (Chen et al. 2020). Hence, it is plausible that CD126-negative tTregs may exhibit superior functional activity (Chen et al. 2020). Secondly, iTregs, but not tTregs, exhibit negligible levels of suppressor of cytokine signaling 1 (SOCS1) and suppressor of cytokine signaling 3 (SOCS3) in response to IL-6 stimulation (Saleh et al. 2020). SOCS1 has been extensively implicated in mediating Th17 cell differentiation (Knosp et al. 2013). Therefore, the differential expression of SOCS proteins may contribute to the activation of STAT-3 in tTreg cells and the potential conversion of tTreg cells into Th17 cells (Takahashi et al. 2011). Thirdly, it has been observed that TGF-β can upregulate the expression of Bcl-2 and diminish T cell apoptosis in recovered iTregs, indicating that iTregs may exhibit reduced susceptibility to apoptosis compared to tTregs (Wan and Flavell 2008).

DCs represent a pivotal cell subset responsible for initiating immune responses, with their pro- or anti-inflammatory polarization influenced by various environmental cues. Prior investigations have indicated that iTreg cells possess the capability to confer immunoregulatory properties upon DCs (Semitekolou et al. 2018). MS constitutes an autoimmune disease characterized by immune dysregulation, culminating in the infiltration of immune cells into the CNS, thereby instigating demyelination, axonal injury, and neurodegeneration (Attfield et al. 2022). Primarily mediated by myelin-specific autoreactive CD4⁺T cells, the pathogenesis of MS also involves significant contributions from DCs and Tregs (Attfield et al. 2022). Our research has unveiled a novel biological role for iTregs in directing the differentiation of profoundly tolerogenic DCs. This significant finding holds promise for the development of new immunotherapeutic strategies targeting a spectrum of autoimmune disorders.

Our findings demonstrate that iTreg cells function, in part, by modulating DCs, thereby orchestrating the delicate balance between immunity and tolerance, consequently contributing to the amelioration of EAE. In patients with MS or in the EAE murine model, DCs exhibit an activated phenotype and promote the differentiation of T cells into pathogenic Th17 cells. Subsequently, these Th17 cells migrate to the CNS where they instigate attacks on oligodendrocytes, leading to demyelination (Rodríguez Murúa et al. 2022). DC-derived IL-1β and IL-23 further consolidate a pathogenic phenotype in Th17 cells within the CNS. In our study, administration of Treg cells not only inhibits the activation of splenic DCs but also confers tolerogenic activities upon them. This tolerogenic capacity relies on DCs establishing a complex network of cell-to-cell interactions, encompassing direct contact and the release of soluble factors (Li et al. 2022). Various factors contribute to the induction of a tolerogenic DC phenotype, including IL-10, TGF-β, and vitamin D (Domogalla et al. 2017). Studies have revealed the mechanisms by which Treg cells confer tolerogenic attributes to DCs; the most critical signaling pathways identified to date encompass IL-10, CTLA-4, and TGFβ (Domogalla et al. 2017; Morante-Palacios et al. 2021). Notably, the effects mediated by IL-10 were observed predominantly when immature DCs were exposed to IL-10, while mature DCs remained insensitive to IL-10 stimulation, maintaining a stable, mature phenotype (Kim and Kim 2019). In asthmatic mice, Treg cells secrete IL-10, mediating the induction of tol-DCs, and blockade of IL-10 leading to suboptimal generation of tol-DCs (Briceno Noriega and Savelkoul 2021). However, in lupus mice, IL-10 appears dispensable for tol-DC induction. TGF-β, a pleiotropic cytokine known for its role in converting naïve T cells into iTreg cells, exerts suppressive functions by inducing Foxp3 gene and protein expression in T cells. Administration of iTreg cells efficiently suppresses lupus-like chronic graft-versus-host disease by preventing the expansion of immunogenic DCs and inducing tolerogenic DCs (Lan et al. 2012). Protective effects of transferred iTreg cells require both IL-10 and TGF-β. However, the influence of iTregs on the induction of tol-DCs in the EAE model remains unexplored.

Understanding the mechanisms underlying the induction of tolerogenic functional transitions in DCs by Tregs during immune-mediated processes holds promise for the development of new therapeutic strategies or drug targets. The AKT/mTOR signaling axis plays a pivotal role in modulating the maturation, activation, and survival of DCs (Na et al. 2020). mTOR regulates protein synthesis by directly phosphorylating and inactivating the repressor of mRNA translation, eukaryotic initiation factor 4E-binding protein 1 (4E-BP1), and by phosphorylating and activating S6 kinase (p70S6K) (Weichhart and Säemann 2008). The role of mTOR signaling in DC function during inflammatory immune responses remains contentious. Mainstream research indicates that inhibition of the Akt/mTOR pathway, particularly through rapamycin, enhances Treg induction and diminishes the immunological effects of DCs (Weichhart and Säemann 2008; Chou et al. 2022). Additionally, inhibition of AKT and p70S6K phosphorylation facilitates the differentiation of iTreg cells (Zhang et al. 2019). Notably, 1,25(OH)₂D3 has emerged as a significant regulator of the immune system, exerting its effects by inducing immune tolerance in DCs. Treatment with 1,25(OH)₂D3 suppresses the maturation of bone marrow-derived DCs and promotes a dominant tolerogenic function, achieved through inhibition of the AKT/mTOR signaling pathway (Cho et al. 2022). However, from a cellular metabolism standpoint, prior research has underscored the critical role of the PI3K/Akt/mTOR pathway in maintaining the tolerogenic phenotype of 1,25(OH)₂D3-modulated DCs (Ferreira et al. 2015). Our findings corroborate that iTregs suppress the Akt/mTOR pathway, contributing to the induction of an anti-inflammatory profile in murine DCs. Notably, the expression of the FoxP3 gene, a classical Treg marker, is regulated by the HIF-1α/mTOR pathway, underscoring the significance of this signaling cascade in Treg function.

Excessive intake of salt enhances the differentiation of Th17 cells, leading to the development of a highly pathogenic phenotype that exacerbates EAE (Kleinewietfeld et al. 2013). However, contrary to this effect, a prior study has revealed that the function of myeloid DCs remains largely unaffected by salt in vitro. Furthermore, high salt intake exacerbates neuroinflammation in EAE mice independently of mature DCs (Jörg et al. 2016). This indicates that distinct subsets of immune cells exhibit differential responses to NaCl. Notably, the induction of a pro-inflammatory environment by salt appears to involve specific effects on immune cells rather than non-specific activation of all lymphocytes and APCs (Zhang et al. 2015). Developing clinically viable cellular therapies requires consideration of real-world inflammatory challenges. The successful amelioration of disease by iTregs transferred before symptom onset highlights their potential for pre-emptive therapy in clinical practice. While this pre-symptomatic intervention model robustly demonstrates the inherent superiority and environmental resilience of iTregs, we acknowledge that its translational impact would be further expanded by evaluating efficacy in a therapeutic setting after disease onset, which more closely mirrors clinical intervention in established autoimmunity. Consequently, beyond their established mechanistic prowess, iTregs represent an innovative and adaptable platform for next-generation immunotherapy, meriting further investigation into their ability to reverse ongoing neuroinflammation.

Conclusion

In summary, our findings offer further evidence supporting the notion that iTregs exhibit distinct biological properties compared to tTregs. Also, these results underscore the potential clinical relevance of iTregs in patients diagnosed with autoimmune and inflammatory conditions, specifically highlighting the importance of considering the complex influence of environmental factors such as diet.

Abbreviations

iTregs

Induced CD4 ⁺Foxp3 ⁺ regulatory T cells

tTregs

Thymus-derived CD4 ⁺ CD25 ⁺ regulatory T cells

Tregs

Both Treg subsets (tTregs and iTregs)

CFSE

Carboxyfluorescein diacetate succinimidyl ester

EAE

Experimental autoimmune encephalomyelitis

Authors’ contributions

Song Guo Zheng and Yang Luo: Conceptualization, Methodology, Funding acquisition, Figure Supervision, Writing, and Editing. Yang Luo and Jiale Tian: Conceptualization, Funding acquisition, Investigation, Data curation, Software, Writing – original draft. Yong Wang and Haolin Li: Investigation, Resources, Writing – original draft. Xiaofeng Wei and Yating Li: Conceptualization, Writing – editing, Software. Long Zhang: Investigation, Resources, Funding acquisition. David Brand: Investigation, Resources. All authors read and approved the final draft.

Funding

The National Natural Science Foundation of China (81960293, 81871224, 82371817); the Natural Science Foundation of Gansu Province (20JR5RA3); the Joint Research Fund of Gansu Province (23JRRA1495), and the China Postdoctoral Foundation project (2023M731460). the Lanzhou Chengguan District talent innovation and entrepreneurship project (2023RCCX0021), the Hui-Chun Chin and Tsung-Dao Lee Chinese Undergraduate Research Endowment (LZU-JZH2634) and the First Hospital of Lanzhou University excellent doctoral research start-up fund (ldyyyn2018-23).

Data availability

All data generated or analysed during this study are included in this article. Further enquiries can be directed to the corresponding author.

Declarations

Ethics approval and consent to participate

I confirm that I have read the Editorial Policy pages. All animal experiments in this study were approved by the ethics committee of the first hospital of Lanzhou university (protocol code: LDYYLL-2023–454). All applicable international, national, and/or institutional guidelines for the care and use of animals were followed.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.