Background

The immune system is a complex network of cells and molecules that defends the body against harmful pathogens while maintaining self-tolerance. This delicate balance is orchestrated by various cell types, including T cells, which play a central role in adaptive immunity. Regulatory T cells (Tregs) are a specialized subset of T lymphocytes that play a pivotal role in maintaining immune system homeostasis and preventing excessive immune responses such as autoimmune diseases. More than four decades ago, Tregs emerged as a cornerstone of immunological research. Tregs encompass a heterogeneous population of cells with varying origins and functions. Functionally, Tregs constitute the physiological counterplayers of conventional or cytotoxic T cells and crucially contribute to the maintenance of peripheral immune tolerance [1]. Since their identification by Sakaguchi et al. in 1995, regulatory T cells have grown into a large and complex family of regulatory cell populations. Among these, thymically derived Tregs (tTregs), which develop within the thymus before being released into the periphery, represent the majority of peripheral FoxP3 + Tregs. In contrast, peripherally induced Tregs (pTregs) develop from mature conventional FoxP3−CD4+ T cells upon continuous antigen stimulation in peripheral tissues. During this process, conventional T cells acquire regulatory properties directed by multiple factors, including the presence of certain cytokines and the formation of cellular synapses between various immune cells. This duality highlights the dynamic nature of Treg development and its adaptability to various immunological contexts. Within this review we will summarize biology and functions of Tregs and present the current understand of Tregs deficiencies in monogenetic immunodeficiency and multifactorial autoimmune diseases. Additionally, we will discuss novel therapeutic approaches using Tregs as target or agent to overcome currently unmet medical needs.

Treg biology and function

The functional characteristics of tTregs and pTregs overlap but differ in terms of their stability. pTregs show a high plasticity and exert regulatory functions only temporarily by transient expression of FoxP3 and additional regulatory elements, which induce the formation of regulatory cytokines [2, 3]. In contrast, tTregs express high levels of FoxP3 [4] and IL2R alpha chain CD25, but low levels of IL-7 receptor CD127. These characteristic elements are critical for the development, function, and homeostasis of tTregs and are tightly linked to their regulatory stability irrespective of the immunologic milieu [5, 6]. However, despite substantial efforts and the discussion of various promising candidates, a phenotypic marker or marker combination that is uniquely expressed by tTregs or allows the discrimination between tTregs and pTregs has not yet been identified [7].

In humans, Tregs constitute only 3–10% of the naïve peripheral CD4+ T-cell population. During embryogenesis, Tregs are present within the thymus at 12 gestational weeks and remain stable throughout pregnancy and infancy [8]. Fetal tTregs already express FoxP3 and other markers characteristically linked to their early established immunosuppressive phenotype, e. g. the cytotoxic T-lymphocyte-associated protein 4 (CTLA-4) and glucocorticoid-induced TNFR-related protein (GITR) [8, 9]. To protect the human body from autoimmunity, tTregs possess a T-cell receptor (TCR) with a specific affinity for autoantigens [10]. TCR-dependent maturation is mediated by the thymic selection process of tTregs, which focuses on self-protection through the presentation of autoantigens. The presentation of various self-peptides, the so-called tissue-specific antigens (TSA), in medullary thymic epithelial cells (mTECs) is regulated by the transcription factor AIRE (Autoimmune Regulator) and the zinc finger protein Fezf2 [11, 12]. T-cell selection and maturation in a TSA-rich environment ensures immunological self-tolerance. Only T cells bearing TCRs with an intermediate affinity for self-peptides differentiate into tTregs. In contrast, T cells are deleted if they recognize self-peptides with a high-affinity TCR or differentiate into naïve CD4+ T cells if self-peptides are recognized with low-affinity TCRs [7, 13,14,145], which acts as a scaffold bridging the CD28 to CARD11 and NFκB signaling cascades [146]. Functional analysis of CD4+ and CD8+ T-cell responses confirmed deficient CD3/CD28 costimulation in CARMIL2 deficient individuals [143]. While peripheral T-, B-, and NK cell counts are typically normal, Tregs are profoundly reduced. Due to deficient T-cell maturation, both CD4+ and CD8+ T-cell subsets are skewed towards naïve forms [140,141,142,143]. Within the B-cell compartment, class-switched B cells and plasmablasts may be reduced and show impaired immunoglobulin formation [141, 142]. Clinically, this results in a combined immunodeficiency syndrome with profound, early onset skin and inflammatory bowel disease and susceptibility to infections [147,148,149].

BACH2 deficiency

The transcription factor BACH2, a highly conserved basic leucine zipper protein, is a key modulator of multiple immune processes, including T- and B-cell differentiation and maturation [150,151,152]. The BACH2 locus contains a T-cell super-enhancer that regulates the expression of multiple pro-inflammatory cytokines and cytokine receptors [153, 154] and thereby reducing effector T-cell differentiation. In Tregs, BACH2 induces high FoxP3 expression, thereby promoting Treg development, maturation, and survival [151, 155]. BACH2 haploinsufficiency causes low Treg frequency and function, while differentiation of Th1-cells, which express the intestinal homing receptors CCR9 and ITGB7, is enhanced [97]. Similarly, the lack of BACH2 mediated repression of Th-2 differentiation results in increased Th-2 cytokine formation, promoting both airway and bowel inflammation [156]. As the effects of BACH2 deficiency manifest at every level of B-cell development, B-cell maturation and IgG class switch are profoundly impaired, resulting in increased transitional B-cell numbers, low immunoglobulins, and inability to generate appropriate antibody responses to specific antigens of vaccines. Accordingly, the clinical picture of BACH2-related immunodeficiency and autoimmunity (BRIDA syndrome) syndrome is dominated by sinopulmonary infections and autoimmune gastrointestinal diseases, which may present early in life [154].

Tregs in autoimmune diseases

Unlike monogenetic Treg disorders polygenetic or multifactorial Treg deficiencies involve a complex interplay of multiple genes and environmental factors. Due to the multifactorial and polygenetic nature of these diseases understanding the interwoven factors contributing to the specific pathophysiology remains challenging. As mentioned earlier, we now know that Tregs represent a diverse subpopulation characterized by distinct transcriptional repertoires influenced by tissue- or context-specific transcription factors. For example, Tregs residing in adipose tissue express the transcription factor PPARγ, whereas those critical for driving Th1-type responses increase Tbet [58, 157, 158]. However, our current challenge is to use this knowledge to identify biomarkers that indicate Treg function in clinical settings. The broad spectrum of Treg functions makes the selection of a single marker or in vitro functional assay challenging, particularly in the context of a particular disease. The difficulties become even greater when assessing Treg activities in humans, primarily because of the obstacles associated with isolating Tregs from tissues other than blood.

Tregs in type 1 diabetes

Type 1 diabetes (T1D) is the best-characterized autoimmune disease, and is often referred to as Juvenile Diabetes. It is a persistent autoimmune ailment characterized by a targeted immune response driven by both T- and B-cells, culminating in the destruction of insulin-producing β-cells nestled in the pancreatic islets [159]. T1D is one of the most common chronic metabolic diseases, affecting approximately 1.5 million people under 20 years of age [160]. It is one of the most frequent chronic metabolic diseases in childhood and adolescence, with a global increase in the incidence rate of 3–4% per year and strong regional differences [161]. Many autoimmune disorders, including T1D, frequently share disruptions in the control of effector cell populations as a fundamental contributing element [162, 163], and this aberration might stem from irregularities in the suppressive functions governed by Tregs.

A significant number of studies have indicated no disparities in peripheral blood Treg frequencies among T1D patients [164]. Nonetheless, anomalies in Treg phenotype and their suppressive potential have been documented [165, 166]. As described above, the challenge of obtaining healthy human tissue is particularly daunting when studying the role and function of Tregs in T1D, as pancreatic samples can only be obtained postmortem. Unfortunately, the unavailability of pancreatic samples from T1D patients has primarily confined data collection to peripheral blood, obscuring whether Tregs actively mitigate β-cell destruction or exhibit modified traits within islets during disease progression. Consequently, animal models such as mice have been harnessed to scrutinize disease advancement within the islet microenvironment.

Therefore, non-obese diabetic (NOD) mice are an essential model for T1D research. NOD mice spontaneously develop autoimmune diabetes, typically commencing at approximately 12 weeks in females, with the incidence increasing until approximately 25 weeks [167]. Male NOD mice experience delayed onset and progression of diabetes. The incidence is approximately 70% in females and 30% in males, a difference potentially rooted in gender-based variances in the gut microbiome and hormonal fluctuations [168]. Environmental factors, including housing conditions and diet, have been implicated in autoimmune diabetes onset [167]. Genomic investigations have identified susceptibility loci termed as insulin-dependent diabetes (IDD) loci in NOD mice. A plethora of over 40 IDD loci has been cataloged, with the major histocompatibility complex (MHC) exhibiting the most substantial link to T1D incidence [167, 169, 170]. Although the NOD mouse manifests several similarities with human T1D, some distinctions persist. Nevertheless, NOD mice have emerged as valuable tools for elucidating the role of Tregs in autoimmune diabetes [167].

Undoubtedly, genetic susceptibility constitutes a fundamental cornerstone in the evolution of T1D, with a significant proportion of susceptible single nucleotide polymorphisms (SNPs) being closely linked to immune-related genes, thereby underscoring immune dysregulation. Particularly noteworthy is the robust correlation observed with genes that have considerable influence over Treg function, most prominently IL2RA, IL-2, PTPN2, CTLA4, and IL-10 [171, 172]. However, translation of these SNPs into functional outcomes has only been achieved in a few studies. Additionally, because numerous pivotal genes serve both effector T cell and Treg functions, deciphering the relative impact of allelic variants on regulatory and effector T cells poses a formidable challenge. Several studies have reported SNP-associated impairments in Treg function, with a particular emphasis on IL-2 signaling [173,174,175]. These findings, coupled with analogous findings in NOD mice highlighting a deficiency in IL-2 signaling within Tregs, have galvanized efforts to harness this pathway for therapeutic intervention [175].

A regrettable adverse facet of low-dose IL-2 therapy, which effectively amplified Tregs, was the simultaneous escalation of eosinophils and natural killer cells, coupled with a reduction in C-peptide levels [176]. However, recent studies on Treg-specific IL-2 administration hold promise for overcoming off-target effects [177,178,179]. Furthermore, novel methodologies geared towards manipulating the pharmacokinetics of IL-2 therapy are expected to enhance its efficacy. One notable study employed the administration of low-dose IL-2/CD25 fusion protein, forestalling diabetes onset and even managing overt diabetes in the NOD mouse model of T1D. The augmented half-life of this IL-2 analog facilitates prolonged interaction with CD25-expressing Tregs, thereby amplifying IL-10 production and encouraging its migration to the pancreas [180].

More recently, a study harnessed T-cell population-specific epigenetic analysis to precisely locate susceptible SNPs within enhancer regions pivotal for Treg function in autoimmunity [181]. Comparative epigenetic evaluations across Treg and conventional T-cell populations revealed that autoimmune-associated SNPs were enriched in naïve Treg-specific demethylated regions and, to a lesser extent, in activated Treg-demethylated regions. These insights suggest that autoimmune-linked SNPs exert a more profound influence on thymus-derived Treg development and function than on aberrant activation of autoimmune effector T cells.

Several pathways critical for Treg development, function, and lineage stability are perturbed in T1D, potentially resulting in Treg dysfunction. Although studies employing Tregs derived from peripheral blood have provided evidence of altered Treg function in T1D patients [182, 183], the extent to which peripheral blood can accurately reflect Treg function at the tissue site remains ambiguous. Mouse models, particularly the NOD model, have provided invaluable insights into the mechanisms underlying the Treg suppression of islet autoimmunity. These investigations have illuminated the notion that certain deficiencies in Treg function are exacerbated at the tissue site, with Treg deficits not always conspicuous in in vitro assays [184, 185]. It is plausible that an amalgamation of chronic inflammatory mediators, anomalies in the IL-2 signaling pathway, and diminished TCR diversity, among other factors, converge within the pancreatic tissue, collectively weakening Treg function [184,185,186,187]. An optimal therapeutic strategy tailored to Tregs should be meticulously devised to address this combination of defects by stabilizing FOXP3 [188, 189].

Tregs in autoimmune hepatitis

The incidence of autoimmune hepatitis (AIH) is similar. Due to the much smaller number of patients, the data were more uncertain. A pioneering genome-wide association study (GWAS) identified mainly genes of the HLA complex as risk factors, such as in T1D, but none were directly related to Tregs [190]. In relation to AIH, research has shown that Tregs may play a significant role in this context. A couple of studies from Vergani et al. have observed that patients with AIH may have reduced Treg numbers in the peripheral blood compared to healthy individuals [191,192,193]. The reduction in Tregs may contribute to an uncontrolled immune response against endogenous liver cells, ultimately leading to inflammation and liver damage. In particular, in pediatric AIH, decreased Treg, characterized as CD3+CD4+CD25+ numbers and impaired Treg function have been documented. Nonetheless, in a more recent study that included FOXP3 in Treg characterization (CD3+CD4+CD25highFOXP3+), the opposite was described as patients with AIH had increased Treg numbers in the blood [194]. This is consistent with the intrahepatic observation that the number of Tregs increases during active disease [195]. Notably, the same group also observed the opposite result in untreated pediatric patients with AIH [196]. In addition, the more pronounced effect of standard steroid therapy on decreasing Tregs over T effector cells was striking, leading to an increase in the apoptosis of Tregs [195, 197].

In mice, knockout of genes related to Treg development or function, such as aire or pd-1, leads to fatal autoimmunity in the liver and other organs, accompanied by decreased or absent numbers of Tregs [198,199,200]. In contrast, in animal models of AIH that resemble different aspects of human disease, an intrahepatic increase in Tregs in active AIH has also been observed [201,202,203,204].

Tregs in colitis

The role of Tregs in colitis is closely linked to the balance between the inflammatory and regulatory immune responses in the gut. Studies have shown that Tregs play a key role in preventing excessive inflammatory processes in the gut. Patients with colitis, particularly Crohn's disease (CD) and ulcerative colitis, have been observed to have deficiencies in the number and function of Tregs [205,206,207,208]. This, in turn, promotes an unbridled and chronic inflammatory response that can damage intestinal tissue.

Previously conducted genome-wide association studies have identified over 100 separate genetic loci that contribute to either susceptibility or defense against the development of inflammatory bowel disease (IBD), with a considerable portion of these loci being shared between the two conditions [209, 210]. Administration of NOD2 ligands, including peptides or muramyl dipeptides, has demonstrated the potential to alleviate colitis induced by 2,4,6-trinitrobenzenesulfonic acid (TNBS) or dextran sulfate sodium (DSS) in normal mice [211, 212]. In a TNBS-induced colitis model, treatment with Lactobacillus peptidoglycan increased the number of Tregs in mesenteric lymph nodes and elevated IL-10 expression in the colonic mucosa, implying that NOD2 activity within the intestinal mucosa fosters a milieu conducive to immune tolerance. Furthermore, receptors associated with T cell and Treg migration, such as CD62L, C–C chemokine receptor (CCR)4, CCR5, CCR7, CCR9, αEβ7 integrin, and α4β7 integrin, also contribute to the pathogenesis of IBD [213,214,215,216,217,218,219]. The presence of these receptors on Treg cells plays a pivotal role in maintaining intestinal immunological equilibrium, and their compromised expression has been linked to the development of IBD, owing to the impaired migration of Treg cells into the intestinal tract. For example, the absence of CCR7 impedes Treg cell functionality in an experimental colitis model [214].

Tregs as therapeutic agent or target

Among genetic and multifactorial autoimmune diseases, IBD is the most promising disease for polyclonal Treg transfer therapies. It has been shown, that Tr1 cells inhibit the proliferation of antigen-specific T cells through an IL-10-dependent mechanism and exhibit protective effects in the adoptive transfer model of colitis involving naïve T cells, as in SCID patients with colitis [220]. Although both FOXP3 + Treg cells and Tr1 cells generate IL-10, Tr1 cells appear to play a crucial role in upholding tolerance towards commensal bacteria. Nonetheless, Battaglia et al. demonstrated the necessity of FOXP3 + Treg cells next to Tr1 cells to persist for the initial induction of tolerance in autoinflammatory diseases [221,222,223]. We have highlighted the importance of Tregs to home to the gut and expand into the lamina propria to regain immunological tolerance [224]. As observed in SCID patients, the transfer of polyspecific Tregs is sufficient to treat autoimmune diseases that lack functional Tregs, such as SCID or APS-1 patients. Later, this was shown in other studies in the corresponding aire-deficient mouse model [199, 225,226,227,228]. To ensure the safety and efficacy of poly- or ovalbumin-specific autologous Tregs in the treatment of CD, many clinical trials have been conducted, including NCT03185000 (TRIBUTE) [229, 230], Eudract no. 2006–004712-44 [231], NCT02327221, NCT05566977, NCT03011021, NCT02932826, and NCT02691247. In addition, also antigen-specific Tregs bearing a chimeric antigen receptor (CAR) against model antigens were very successful in controlling colitis in animal models [232,233,234,235].

In T1D, the situation is very different, as polyspecific Tregs show no positive effects in mouse models or patients. It has been shown that antigen-specific Tregs are required to control diabetes and prevent its induction. Considering that Treg insufficiency potentially fuels T1D and autoimmune diabetes, bolstering the Treg count in circulation could serve as a strategy to counter this inadequacy. Notably, recurring adoptive Treg transfers into neonatal NOD mice have demonstrated the ability to postpone the onset of autoimmune diabetes, implying that Treg number or functionality might wane in NOD mice over time, necessitating supplementation [236]. Many T1D studies use BDC-2.5 mice, which are genetically modified NOD mice, carrying a transgenic TCR that recognizes a pancreatic antigen in NOD mice. These T cells destroy the insulin-producing cells in the pancreas. Therefore, another convincing strategy involves the adoptive transfer of a small quantity of DC-expanded BDC2.5 TCR-tg Tregs into pre-diabetic NOD mice. The transfer successfully prevented diabetes development and even salvaged mice with manifest diabetes [237]. When pre-diabetic NOD splenocytes or BDC2.5 TCR-tg Teff cells are transplanted into immunodeficient NOD mice, autoimmune diabetes typically emerges approximately 14 d post-transfer. Interestingly, co-transplantation with over a million polyclonal Tregs or a few thousand BDC2.5 TCR-tg Tregs can prevent the disease [238]. While a minimal number of antigen-specific Tregs have the capacity to reverse autoimmune diabetes, adopting ten-fold more polyclonal Tregs was not as effective in the therapeutic treatment of NOD mice, underscoring the critical significance of specificity for β-cell antigens in optimizing Treg functionality [239,240,241,242,243].

Clinically, in vitro-expanded polyclonal Tregs are being evaluated as a promising avenue that diverges from pharmacologically based treatments. Early phase clinical trials encompassing both pediatric and adult participants with autologous, polyspecific Tregs have been conducted, reflecting no immediate safety concerns, such as ISRCTN06128462, NCT01210664, NCT02932826, and NCT02772679 [244,245,246,247]. Notably, in children, potential efficacy has been assessed based on C-peptide levels at 4–5 weeks post-treatment. However, while initial elevations in C-peptide levels were evident at the one- and two-year follow-ups, they gradually diminished over time. Intriguingly, nearly 25% of transferred Tregs, characterized by a naïve/memory-like profile, persisted in patients at the one-year follow-up based on deuterium incorporation. A parallel trial conducted in Poland has yielded encouraging outcomes. In a cohort of 12 children with T1D, a one-year follow-up revealed augmented C-peptide levels and reduced insulin usage in 8 of 12 patients, resulting in complete insulin independence in 2 of the 12 patients [244,245,246,247]. Whether these encouraging observations endure and can be replicated in phase 2 clinical trials remains to be ascertained.

The potential for a more robust success may rely on combination therapy. Potential synergies with Tregs have been explored to optimize their therapeutic response in different autoimmune diseases. One effective strategy could involve bolstering the Treg population through the infusion of ex vivo expanded Tregs, while concurrently reducing the Teff population using agents such as anti-CD3 monoclonal antibody (NCT00129259) [248] or LFA3-Immunglobulin (Ig) (NCT00965458) [249, 250], which have shown promise in initial trials for new-onset T1D, followed by more than 30 other trials with similar results.

Another avenue to consider is the coupling of Treg cell infusion with interleukin-2 (IL-2). In vivo, IL-2 at low doses plays a pivotal role in the growth and survival of Tregs [251, 252], and constitute a critical component of Treg expansion protocols. Either Tregs might be directly equipped with IL-2 signals [253] or employing low-dose IL-2 has been effective in enhancing endogenous Tregs, leading to diabetes prevention and reversal in the NOD mouse model [184, 254]. Preliminary clinical investigations involving low-dose IL-2 have demonstrated selective increases in Tregs, yielding positive clinical outcomes in many (auto-)inflammatory conditions [255,256,257,258,259]. While a clinical trial combining IL-2 with rapamycin showed transient deterioration in beta cell function, potentially due to the relatively higher IL-2 dose or the influence of rapamycin, which led to significant increases in natural killer cells and eosinophils, early studies on lower IL-2 doses in T1D have shown no acute degradation in beta cell function [260, 261]. Several ongoing studies are aimed at evaluating the safety and efficacy of this approach. In an upcoming phase I trial involving autologous ex vivo expanded Tregs followed by low-dose IL-2, we will assess whether low-dose IL-2 enhances in vivo survival and functionality of infused Tregs.

Tregs administered in the initial clinical trials on T1D did not exhibit TCRs or CARs that specifically targeted diabetes antigens. There are several possible explanations for the selection of polyclonal Tregs for such an administration. The safety aspect was the most important, as all studies validated the safety of polyclonal cells, which is a primary consideration in devising a clinical research protocol. Thus, the effectiveness of polyspecific Tregs has been demonstrated in animal models, and data from animal studies involving exogenous IL-2, which augments Treg numbers, suggest the efficacy of polyclonal cells. Unlike other cellular processes, Treg-mediated suppression lacks antigen specificity; hence, polyclonal T cells should be capable of regulating cells that possess specificities for diabetes antigens.

Preclinical investigations have proposed that antigen-specific Tregs are more efficient in controlling autoimmune-mediated beta cell destruction than polyclonal Tregs [239,240,241,242,243]. Although target antigens for T1D have been identified, a significant challenge lies in isolating these less common cells from peripheral circulation and subsequently expanding them for clinical application [262]. Hence, it may be advantageous to generate antigen-specific de novo Tregs. Engineered Tregs featuring CARs have exhibited success in pre-clinical T1D [263, 264] and could potentially be viable for clinical deployment, albeit the specific antigen profile might need customization for each individual patient.

Polyspecific Tregs were also used in a clinical trials for AIH (NCT02704338) [265], Pemphigus Vulgaris NCT03239470, and Lupus Erythematosus NCT02428309 proving safety. However, polyspecific Treg transfer was not very effective in AIH mouse models [201, 203]. Therefore, we and others have successfully used optimization protocols for Tregs to determine their therapeutic responses in different AIH models [266,267,268]. Nonetheless, data on combination therapy or antigen-specific CAR-Tregs are lacking.

Notably, the significance of antigen-specific CAR-Treg cells in preventing genetic resilience against organ-specific autoimmunity and inhibiting autoimmune tissue damage has been well documented in various disease models, including multiple sclerosis [269] and Alzheimer’s disease [270].

Conclusion

Numeric and functional Treg deficiencies may be due to monogenetic, inborn errors of immunity or occur by a more complex multifactorial processes. While congenital global Treg deficiencies result in polytopic diseases and usually affect multiple organ systems, genetically undefined autoimmune diseases, such as T1D or AIH, are characterized by the lack of particular antigen-specific Treg subsets but normal Tregs numbers. With the rapidly growing understanding of the genetic and functional Treg biology, new therapeutic approaches are currently being developed to overcome unmet medical needs in Treg diseases. The adoptive transfer of ex vivo expanded or even genetically modified Treg products is in the focus of upcoming clinical trials. The success of these approaches will depend on both overcoming technical and regulatory hurdles and ensuring product safety and efficacy in larger patient cohorts. Undoubtedly, the possibility to tailor highly personalized Treg products that address patient or disease specific needs has certainly opened a new dimension of target specific treatment approaches.