Intestinal Regulatory T Cells

Abstract

Mucosal surfaces are distinctive sites exposed to environmental, dietary, and microbial antigens. Particularly in the gut, the host continuously actively adapts via complex interactions between the microbiota and dietary compounds and immune and other tissue cells. Regulatory T cells (Tregs) are critical for tuning the intestinal immune response to self- and non-self-antigens in the intestine. Its importance in intestinal homeostasis is illustrated by the onset of overt inflammation caused by deficiency in Treg generation, function, or stability in the gut. A substantial imbalance in Tregs has been observed in intestinal tissue during pathogenic conditions, when a tightly regulated and equilibrated system becomes dysregulated and leads to unimpeded and chronic immune responses. In this chapter, we compile and critically discuss the current knowledge on the key factors that promote Treg-mediated tolerance in the gut, such as those involved in intestinal Treg differentiation, specificity and suppressive function, and their immunophenotype during health and disease. We also discuss the current state of knowledge on Treg dysregulation in human intestine during pathological states such as inflammatory bowel disease (IBD), necrotizing enterocolitis (NEC), graft-versus-host disease (GVHD), and colorectal cancer (CRC), and how that knowledge is guiding development of Treg-targeted therapies to treat or prevent intestinal disorders.

9.1. Introduction to Intestinal Tregs

Intestine is a unique site that is constantly exposed to wide array of luminal microbial and dietary antigens. Consequently, while safeguarding the host against the harmful pathogens, the intestinal mucosal immune system has to remain resilient and tolerant to the innocuous food antigens and commensal microbes to be able to endure the immunological insults without getting unduly provoked. The intestine employs many defense mechanisms to maintain its immunosuppressive tone, such as physical barrier (mucus and epithelial layers) or the production of secretory IgA. Tregs are another instrumental players in keeping a check on the activation of local T cells in the intestine (Powrie and Mason 1990; Powrie et al. 1993; Gambineri et al. 2003; Josefowicz et al. 2012).

To better understand the roles of Tregs in the gut, it is important to place it in the context of the intestinal tissue architecture. Histologically, along its entire length, intestine is made up of four layers: mucosa (further consists of the epithelial layer, lamina propria, and muscularis mucosae), submucosa, muscularis propria, and serosa. Lamina propria (LP) compartment consists of a variety of immune cells such as plasma cells, macrophages, dendritic cells, mast cells, and B and T lymphocytes, including Tregs, the subject of this overview. Small intestine (SI) comprises of duodenum, jejunum, and ileum and is characterized by the highly absorptive surface created by finger-like projections called the villi, further enlarged by cellular microvilli located at the apical, lumenfacing membrane of the enterocytes. The small intestinal epithelium consists of invagination forming the crypts of Lieberkuhn. At the base of the crypts reside the intestinal stem cells, which give rise to transient proliferative cells that differentiate and mature into Paneth cells, goblet cells, enteroendocrine cells, and tuft cells as they travel up through the transition zone, to ultimately shed into the lumen at the apex of the villi. Peyer’s patches are organized lymphoid structures unique to the small intestine. They harbor immune cells and are separated from the lumen by a follicle-associated epithelium which contains specialized cells called microfold cells (M cells) capable of sampling the antigens directly from the lumen and delivering it to antigen-presenting cells juxtapositioned at their basolateral side (Fig. 9.1a). The large intestine consists of cecum, colon, and rectum and harbors the bulk of the commensals. Its major function is to absorb water and vitamins while converting undigested food into feces. The most notable histological differences are lack of villi, Paneth cells (except for their metaplastic occurrence in chronic inflammatory bowel diseases), and Peyer’s patches (Fig. 9.1b).

Fig. 9.1.

Features of the small intestinal and colonic mucosa. (a) Small intestinal epithelium. Enterocyte-derived antimicrobial peptides: RegIIIγ, β-defensins, cathelicidin. Paneth cell-derived antimicrobial peptides [lysozyme, α-defensins, secretory phospholipase 2 (sPLA2)]. M cells and goblet cells facilitate luminal antigen transfer to the underlying dendritic cells. (b) Colonic epithelium. Goblet cells produce two-layered mucus layer through Muc2 mucin secretion along with other proteins contributing to the colonic epithelial barrier such as anterior gradient protein 2 homolog (Agr2), small lectin-like protein ZG16 (zymogen granulae protein 16), secreted goblet cell-derived protein Clca1 (chloride channel regulator, calciumactivated-1), or Other abbreviations: DC dendritic cell, IEL intraepithelial lymphocyte. (Figure adapted from Allaire et al. with permission) (Allaire et al. 2018)

Most of the information available regarding the development of intestinal Tregs is based on research on mice. In other organs, Tregs comprise approximately 10% of total CD4⁺ T-cell population However, in the gut, they represent a much higher fraction with >30% of CD4⁺ T cells in the colonic lamina propria (cLP) and about 20% in the small intestinal lamina propria (siLP) (Weiss et al. 2012; Round and Mazmanian 2010; Atarashi et al. 2011; Geuking et al. 2011; Stefka et al. 2014). The presence of higher proportion of Tregs in the colon, the site of the highest microbial biomass, intuitively suggests that development and maturation of intestinal Tregs is strongly influenced by microbial antigens. Indeed, germ-free (GF) mice and long-term broad spectrum antibiotic-treated specific pathogen-free (SPF) mice have a dramatic reduction in their colonic Tregs compared to their respective controls (Atarashi et al. 2011; Geuking et al. 2011; Weiss et al. 2012). However, the comparable numbers of Tregs present in the siLP of GF and SPF mice indicate that in this intestinal segment, bacterial antigens contribute to the Treg population to a lesser degree (Kim et al. 2016). The same study showed that once the weaned GF mice were switched to antigen-free diet, siLP (but not cLP) Tregs declined, thus revealing a significant role of dietary antigens in regulation of small intestinal Treg population (Kim et al. 2016).

Human CD4⁺ Tregs are characterized by expression of the transcription factor forkhead box P3 protein (FOXP3, scurfin), high surface expression of CD25, and low or no expression of CD127. Tregs express high levels of CD25 (the IL-2 receptor α-chain) due to their high dependence on interleukin 2 (IL-2) for their development (Zheng et al. 2007; Davidson et al. 2007) and maintenance and peripheral homeostasis (D’Cruz and Klein 2005; Fontenot et al. 2005). The groundbreaking study led by Sagakuchi demonstrated that depletion of CD25⁺ subpopulation of CD4⁺ T cells resulted in the manifestation of multiorgan autoimmunity, including intestinal inflammation (Sakaguchi et al. 1995). A few years later, Fiona Powrie and colleagues successfully developed the CD4⁺CD45RB^hi T-cell transfer model of IBD. The consequent discovery that the symptoms of this model could be suppressed via the co-transfer of CD4⁺CD25⁺ T cells, was a huge leap in the field (Mottet et al. 2003). Later, FoxP3 was discovered to be a vital transcription factor for the induction and maintenance of CD4⁺CD25⁺ Tregs. In “scurfy” mice, an X-linked recessive mutation led to scaly skin, lymphoproliferation, hypergammaglobulinemia, lymphadenomegaly, anemia, runting, and early death (Godfrey et al. 1991). In 2001, seminal genetic studies in the “scurfy” mice led to the discovery of Foxp3 gene (Brunkow et al. 2001). Other contemporary studies revealed that human patients with IPEX syndrome (immunodysregulation polyendocrinopathy enteropathy X-linked syndrome) harbored mutations in Foxp3 gene and its 3′UTR region (Bennett et al. 2001; Wildin et al. 2001). Conspicuously, the FoxP3-deficient mice and patients with IPEX syndrome present with gastrointestinal manifestations in the form of autoimmune enteropathy (Powrie et al. 1993; Gambineri et al. 2003), thus highlighting the role of Tregs in the maintenance of intestinal homeostasis.

While CD4⁺ Tregs are most widely studied in health and disease, CD8⁺ Tregs were identified in 1970 by Gershon and Kondo, long before the characterization of CD4⁺ Tregs (Gershon and Kondo 1970). CD8⁺ Tregs are a heterogenous population that are not very well characterized and their role in the maintenance of intestinal homeostasis is poorly understood. In this chapter, we will primarily focus on CD4⁺ Tregs, while the reader is directed to published review articles on CD8⁺ Tregs and original articles therein (Flippe et al. 2019; Yu et al. 2018; Tsai et al. 2011; Liu et al. 2014; Zhang et al. 2019).

9.2. Intestinal Tregs Subsets

Perhaps as an adaptation to the highly specialized and dynamic intestinal environment, Treg population is characterized by a significant heterogeneity. The current knowledge about the different Treg subsets is mostly based on mice as just like any other biomedical research, studying Treg biology in humans is very challenging. While Foxp3 is considered to be the signature marker associated with murine Tregs, it is not so straight-forward in human Tregs. Moreover, unlike in mice, where a FoxP3 reporter mouse lines are available, it is a technical challenge to isolate human Tregs based on FoxP3 which is expressed intracellularly, which precludes isolation of live cells for further functional studies. Even though the presence of Foxp3 is indispensable for the development and function of human Tregs, it is not necessarily synonymous with immunosuppressive functions. There have been many instances when expression of FoxP3 was shown in activated human T cells that lacked the regulatory activity (Allan et al. 2007; Morgan et al. 2005; Wang et al. 2007; Gavin et al. 2006; Tran et al. 2007). Human Tregs are also known to express cytotoxic T lymphocyte-associated antigen 4 (CTLA4) and glucocorticoid-induced TNF receptor (GITR), but just like CD25, they are not very specific to Tregs and can be expressed by effector T cells (T_eff) (Read et al. 2000; Takahashi et al. 2000; McHugh et al. 2002). The CD127 and CD27 markers have been proposed to increase the specificity of Treg identification. CD127 expression is much lower in CD25⁺Foxp3⁺ Tregs than in T_eff (Seddiki et al. 2006), and this marker is useful to identify Tregs in non-inflammatory conditions. Therefore, most of the human studies are conducted on CD4⁺CD25^brightCD127^lo cells isolated from peripheral blood.

In the murine gut, based on transcription factor and surface molecule expression, CD4⁺ Tregs can be subdivided into the following subsets: GATA3⁺Helios⁺(Nrp1⁺), retinoic acid receptor-related orphan receptor γt (RORγt)-expressing RORγt⁺Helios⁻ cells, and RORγt⁻ Nrp1⁻(Helios⁻) Tregs.

9.2.1. Thymic vs. Peripheral Tregs

Based on their origin, Tregs are classified into thymic Tregs (tTregs) which develop in the thymus and into peripheral Treg which are induced from naïve T cells in the periphery (pTreg). Murine tTreg differentiation into Foxp3⁺ Treg is a two-step process: First, future Tregs recognize MHCII-presented self-antigen on thymic epithelial cells via a high-affinity T-cell receptor (TCR) (Stritesky et al. 2012; Xing and Hogquist 2012). In the second step, γ_c cytokines (mainly IL2) and CD28 co-stimulation lead to an induction of FoxP3 expression (Xing and Hogquist 2012; Fontenot et al. 2005; Salomon et al. 2000; Hori et al. 2003; Tai et al. 2005). In contrast, future pTregs exit the thymus as naive CD4⁺ T cells and differentiate into Foxp3⁺ Treg cells in the periphery, upon subsequent recognition of their cognate non self-antigen (Shevach and Thornton 2014; Chen et al. 2003; Yadav et al. 2013) and in the presence of a permissive cytokine microenvironment (Zheng et al. 2002, 2004). The role of cytokines and other factors in the development of Tregs is discussed in Sect. 9.4 of this chapter. One of the big challenges that remains in the Treg field is the paucity of markers (especially cell surface proteins) to distinguish the pTregs from the tTregs. So far, Helios and Neuropilin 1 (Nrp1) have shown to be the two most promising markers expressed by tTregs (Thornton et al. 2010; Yadav et al. 2012; Weiss et al. 2012). However, both of these markers are not exclusive to tTregs. The finding that inflammatory milieu upregulates the expression of both Helios and Nrp1, further complicates their utilization as markers (Gottschalk et al. 2012). While Helios is present in 70% of Tregs in peripheral lymphoid tissues (Thornton et al. 2010), the presence of Nrp1 in humans is still questionable. In humans, both of these markers have not proven to be very reliable. A fraction of human tTregs does not express Helios (Himmel et al. 2013), and Nrp1 expression can be induced in activated humans T_eff (Milpied et al. 2009). Intestinal Tregs are thought to have a dual origin as they could either arise in thymus or develop in the periphery, in response to local microbiota and food antigens. However, a vast majority of Tregs present in the small intestine and colon are Helios⁻Nrp1⁻ (Ohnmacht et al. 2015). In fact, it is the frequency of FoxP3⁺Helios⁻Nrp1⁻ Tregs which is reduced in the colon of GF and antibiotic-treated SPF mice (Atarashi et al. 2011; Geuking et al. 2011; Weiss et al. 2012).

Development of FoxP3⁺ pTregs and tTregs is also regulated by Foxp3 intronic enhancers, termed conserved noncoding sequence 1, 2, and 3(CNS1-3). While CNS3 is instrumental in the initial induction of FoxP3 in tTregs, CNS1 harbors a TGFβ-NFAT response element and a binding site for retinoic acid receptor (RAR) and retinoic X receptor (RXR) heterodimer, which is activated by retinoic acid (Zheng et al. 2010; Xu et al. 2010; Tone et al. 2008). CNS1-deficient mice lack pTregs, while their tTregs differentiation remains unaffected (Zheng et al. 2010). Another study reported that CNS1 deficiency in colonic TCR-transgenic T cells only affected initial, but not late, pTreg cell selection (Nutsch et al. 2016). Nonetheless, CNS1-deficient mice display a dysbiotic microbiota with exacerbated Th2 type pathologies in the lung and colon (Josefowicz et al. 2012; Campbell et al. 2018). pTregs have also been shown to suppress Th1-type responses to food antigens in the gut (Kim et al. 2016). Cumulatively, these studies highlight the key role of pTregs in maintaining tolerance toward luminal antigens.

While pTregs remain the predominating Treg population in colon, it has been suggested that tTregs and pTregs supplement the function of each other, partly by expanding their TCR diversity (Haribhai et al. 2011). Moreover, in the absence of pTregs, tTregs can migrate to intestinal LP and expand locally in response to the cross-reactive microbial antigens (Cebula et al. 2013). Using single-cell and high-throughput sequencing on TCR^mini mice with “limited but diversified TCR,” this study demonstrated that TCR repertoires of colonic tTregs and pTregs have significant overlap (Cebula et al. 2013). A study with mice with MHCII expression restricted to the cortical thymic epithelial cells and thus cannot provide peripheral MHCII-TCR interaction necessary for the induction of pTregs demonstrated that majority of intestinal Tregs were of thymic origin (Korn et al. 2014). The study revealed that there exists a unique “Treg niche” in the intestine, where tTregs migrate during early life and are maintained without the need for TCR stimulation. However, broad spectrum antibiotic treatment abrogated LP tTregs, suggesting that continuous microbial stimuli might be vital for the maintenance of Tregs in the intestinal compartment (Korn et al. 2014). In summary, both pTregs and tTregs are likely to contribute to the maintenance of intestinal homeostasis, although it is challenging to determine which of the two populations plays a major role.

9.2.2. GATA3⁺Helios⁺(Nrp1⁺) Tregs

GATA-binding protein 3 (GATA3), the canonical Th2-specific transcription factor, is expressed by up to 65% of intestinal Tregs and can be induced in CD4⁺Foxp3⁺ Tregs by TCR engagement (Wohlfert et al. 2011). In the same report, human Tregs (CD4⁺CD25^brightCD127^lo) isolated from peripheral blood of healthy donors were also able to dramatically upregulate GATA3 expression upon TCR engagement (Wohlfert et al. 2011). However, this subset has not yet been demonstrated in the human gut. GATA3 has been shown to be important for immunosuppressive function and accumulation of Tregs to the extent that GATA3-deficient Tregs were unable to suppress naïve T-cell transfer-induced colitis in mice (Wang et al. 2011). In the mouse gut, GATA3⁺Helios⁺ tTregs have been also found to express IL33 receptor ST2 and amphiregulin (Schiering et al. 2014). IL33 is produced by intestinal epithelial cells under inflammatory conditions. In fact, high levels of IL33 have been reported in inflamed lesions of inflammatory bowel disease patients (Kobori et al. 2010; Pastorelli et al. 2010; Seidelin et al. 2010). In the same line of thought, in steady state, ST2⁺ FoxP3⁺Tregs showed increased production of IL10 and TGFβ and displayed preferential accumulation in the intestinal LP due to high expression of gut-homing receptors CCR9 and α4β7 (Siede et al. 2016). Furthermore, ST2⁺ Tregs display an activated phenotype and express high levels of OX40 (Schiering et al. 2014), a secondary co-stimulatory immune checkpoint molecule which is also important for the accumulation of Tregs in colon and for Treg-mediated suppression of naïve T-cell transfer-induced colitis (Griseri et al. 2010). It is thus plausible that GATA3⁺ tTregs might be home to the sites of damage in the gut to curb excessive inflammation and mediate tissue repair. However, this hypothesis still needs to be verified.

9.2.3. RORγt⁺Helios⁻ pTregs

Another Treg subset that is abundantly located in the gut expresses the Th17 master transcription factor, retinoic acid receptor-related orphan receptor γt (RORγt). The absence of Helios (Sefik et al. 2015; Ohnmacht et al. 2015) and Nrp1 (Yang et al. 2016) on most of the RORγt⁺ Tregs suggests their peripheral origin. Owing to their similar expression of RORγt, the RORγt⁺Foxp3⁺ Tregs were once considered to be precursors for unstable Tregs that could differentiate into Th17 cells (Zhou et al. 2008; Lochner et al. 2008). However, it was later revealed that, in steady state, RORγt⁺Helios⁻ FoxP3⁺ Tregs comprise about 65% and –35% of total Treg population in the colon and small intestine, respectively (Sefik et al. 2015; Ohnmacht et al. 2015; Yang et al. 2016). Comparable frequencies of Rorγt⁺ Tregs were detected by staining cells from healthy or inflamed (Crohn’s) colon biopsies (Sefik et al. 2015) and among healthy human PBMCs (Duhen et al. 2012), confirming the presence of this subset in humans. Interestingly, there were higher number of circulating IL-17-producing RORγt⁺Foxp3⁺CD4⁺ lymphocytes in IBD patients, and the suppressive ability of these Tregs were reduced by approximately 60% in patients with IBD compared with healthy controls (Ueno et al. 2013).

The plasticity between Th17 cells and RORγt⁺ Tregs is not very well understood. Development of Th17 cells is dependent on IL6, IL1β, and IL23. However, while IL6 and downstream STAT3 signaling are important for the development of intestinal RORγt⁺ Tregs, IL1β and IL23 have not been shown to play a role under homeostatic conditions (Sefik et al. 2015; Xu et al. 2010; Yang et al. 2016). Making use of the transcriptomic and epigenetic profiling, Yang et al. (2016) revealed that RORγt⁺ Foxp3⁺T cells display higher similarity to Tregs than to Th17 cells and that they represent a lineage-stable population with potent suppressive activity. Using fixed TCRβ transgenic mice, it was shown that TCR repertoire of RORγt⁺ Foxp3⁺ Tregs is largely unique compared with other colonic T-cell subsets (Solomon and Hsieh 2016). However, they also observed a subset of Foxp3⁺RORγt⁺ TCRs that are shared with Th17 cells. The same study suggested that the development of mucosal RORγt⁺ Foxp3⁺T cells is dependent on CX3CR1⁺ antigen-presenting cells and can be derived from naïve CD4⁺ T cells in the periphery via a RORγt− Treg intermediate before co-expressing RORγt. Furthermore, RORγt⁺ Foxp3⁺Treg subset is readily lost in GF mice or upon antibiotic treatment, signifying the indispensable role of microbiota in their maintenance (Ohnmacht et al. 2015; Sefik et al. 2015). Interestingly, the differentiation of RORγt⁺ Tregs is also dependent upon constant antigen exposure, as mice deficient in MHCII have decreased frequency of this subset (Ohnmacht et al. 2015). The role of microbiota in the development of RORγt⁺ Tregs would be further discussed later in this chapter. Additionally, c-Maf, a transcription factor which has already been known to be a positive regulator of Th17 cells, has also been found to be highly expressed by RORγt⁺ Foxp3⁺Tregs (Xu et al. 2018). Not only has c-Maf been found to be required for the development of RORγt⁺ Tregs, but is also important for their suppressive ability, as ablation of c-Maf in Tregs leads to spontaneous inflammation in mice that is not observed in mice with RORγt-deficient Tregs (Sefik et al. 2015; Xu et al. 2018; Yang et al. 2016).

9.2.4. RORγt⁻ Nrp1⁻ (Helios⁻) pTregs

Another subset of intestinal Tregs are the CD4⁺Foxp3⁺RORγt⁻ T cells. They are believed to be of peripheral origin due to absence of Helios or Nrp1. These cells represent about 50% of SI LP Tregs and 15% of colonic Tregs (Kim et al. 2016). Since these Tregs are more frequently present in SI, they are induced by dietary antigens, so much, so that they are not affected by the absence of microbiota but disappear in mice fed “antigen-free” diet (Kim et al. 2016). In fact, the absence of this Treg subset leads to increased susceptibility to food allergy in mice (Kim et al. 2016).

9.2.5. Effector Tregs and IL10⁺ Tregs

In humans, based on the expression levels of Foxp3 and protein tyrosine phosphatase isoform A (CD45RA), human peripheral blood Foxp3 +CD4⁺ T cells can be classified into Foxp3^loCD45RA⁺ naive Tregs which after antigenic stimulation differentiate into Foxp3^hiCD45RA⁻ Tregs that represent activated, antigen primed, terminally differentiated and highly suppressive effector Tregs or eTregs. In contrast, Foxp3^loCD45RA⁻ cells do not possess suppressive activity and can even secrete pro-inflammatory cytokines (Miyara et al. 2009). To understand these cells better, some studies have been carried out with mouse CD44^hiCD62L^lowKLRG1⁺FOXP3⁺ cells, thought to represent the murine counterpart of the human eTregs (Liston and Gray 2014). The CD44^hi effector Tregs are most abundant in the colonic LP and depend on constant TCR engagement and subsequent IRF4 expression for their maintenance (Levine et al. 2014). On the other hand, TCR and IRF4 are dispensable for the development of CD44^loCD62L^hi naive Tregs as tamoxifen-induced Treg-specific deletion of TCRα or IRF4 does not adversely affect their numbers (Levine et al. 2014). Transcription factors, IRF4 and Blimp-1, jointly regulate the differentiation, function, and homeostasis of eTregs (Levine et al. 2014; Cretney et al. 2011).

Another important Treg subset described in humans is the FOXP3⁻ type 1 (Tr1) cells that develop exclusively in periphery and secrete copious amounts of IL10 and TGFβ1, intermediate levels of IFNγ and no IL4 (Levings and Roncarolo 2005; Roncarolo et al. 2018). In fact, IL10-producing Tr1 cells were originally identified and isolated from patients who have severe combined immunodeficiency (SCID) and had undergone successful HLA-mismatched bone marrow transplantation. In these studies, antigen-specific Tr1 cells were generated in vitro from both mice and human naïve CD4⁺ T cells (Bacchetta et al. 1994; Groux et al. 1997). The in vitro generated murine Tr1 cells were able to prevent establishment of colitis in SCID mice after CD4⁺CD45RB^hi adoptive T-cell transfer in IL10- and TGFβ-mediated fashion (Groux et al. 1997). Hence, Tr1 cells appear to significantly contribute to maintaining the intestinal tolerance.

9.3. Antigen Specificity of Intestinal Tregs

One of major challenges in the field has been to understand the antigen specificity of intestinal Tregs. For example, are pTregs generated against self-antigens or antigens derived from commensal bacteria? Although some earlier studies have shown that commensal bacteria are not necessary for colonic Treg generation (Min et al. 2007; Round and Mazmanian 2010), generation of Tregs in response to foreign antigens has been established with TCR transgenic models of oral tolerance (Curotto de Lafaille et al. 2008; Sun et al. 2007). Earlier in this chapter, we have described studies with germ-free and antibiotic-treated mice that clearly showed the importance of the gut commensals in driving colonic Treg numbers. Some human gut commensals such as Clostridium clusters IV and XIVa (Atarashi et al. 2011) and Bacteroides fragilis (Round and Mazmanian 2010) have been found to increase the frequency or function of colonic Tregs via protease-resistant capsular polysaccharide (PSA) and a TLR2-dependent mechanism (Round et al. 2011). PSA also induced expression of Foxp3 along with CD39 in naïve human CD4 T cells in vitro while promoting IL-10 secretion (Telesford et al. 2015).

To address TCR specificity, labs developed TCR transgenic mouse lines that respond to antigens derived from commensal bacteria, e.g., flagellin (Cong et al. 2009). However, this approach does not allow for the comparison of the normal in vivo frequency of those TCRs in the Tregs with the effector T cells. The study by Lathrop et al. (2011) analyzed the colonic TCR repertoire and concluded that TCR usage in the colon is very different from that in any other organ and that there was very little overlap in TCRs in Tregs and the effector/memory T cells. This study confirmed that the colonic Treg population is strongly shaped and fine-tuned by the local antigenic milieu and that the effector T cells and Tregs are not necessarily activated by the same antigens. Furthermore, it could be safely inferred that colonic Tregs recognize and suppress the immune response evoked against the antigens from commensal bacteria. In fact, “Limited mice,” which express only a limited TCR repertoire, display shrunken Helios⁻ Tregs and eventually develop loss of tolerance to the gut microbiota and spontaneous development of Th17-driven colitis (Nishio et al. 2015).

9.4. Key Factors Required for the Development and Maintenance of Tregs in the Gut

The intestinal Tregs are very unique and highly specialized due to their role in preventing inflammatory responses to commensal flora, diet, and other innocuous antigens. There are many different mechanisms that are engaged in the induction and maintenance of Tregs in the intestine. While most of these intricate pathways have been elucidated in mice, some attempts have been made to validate them in humans.

9.4.1. Host Factors: Cytokines

As already described in the Introduction to Intestinal Tregs, development of Treg precursors in the thymus requires not only the strong TCR stimulation but also co-stimulation through CD28 and a presence of common-γ-chain cytokines such as IL2 and IL15 (Chinen et al. 2016; Setoguchi et al. 2005). The transcription factor STAT5, which is activated downstream of the IL2 receptor α (CD25), binds to CNS2 in the Foxp3 gene and thus stimulates its expression. In fact, many mouse models mimicking IL2 deficiency have demonstrated IL2 to be a key cytokine for the development and the maintenance of tTregs in the periphery (Papiernik et al. 1998; Almeida et al. 2002; Malek et al. 2002; Setoguchi et al. 2005). Accordingly, CNS2-deficient Tregs are sensitive to low levels of IL2 and eventually lose FoxP3 expression in response to strong TCR activation or pro-inflammatory cytokines, a typical intestinal milieu (Feng et al. 2014; Li et al. 2014). On the other hand, Treg pool can be increased following treatment with agonistic complexes of IL2 or antibody to IL2 (Oldenhove et al. 2009; Wohlfert et al. 2011).

The role of IL2 is central to the development of Tregs as IL2-deficient mice have been known to develop colitis (Sadlack et al. 1993). In general, IL2 potently stimulates T-cell growth and population expansion. However, the function of IL2 signaling for Treg differentiation and maintenance of immune system homeostasis appears to overcome its effects on the effector T cells, since mice deficient in IL-2 or components of the IL-2 receptor, IL-2Ra (CD25) or IL-2Rb (CD122), suffer from an aggressive lymphoproliferative, autoimmune syndrome. Similarly, patients with IL-2Ra deficiency develop immunodeficiency and early onset IBD live-disease (Sharfe et al. 1997). It has been reported that while CD25 activation by IL2 is required for FoxP3⁺ pTreg cell differentiation, survival, and expansion, it remains dispensable for the induction and differentiation of FoxP3⁺ tTreg cells (D’Cruz and Klein 2005; Fontenot et al. 2005). However, a later report showed that IL2 seems to be critical for the maintenance of intestinal GATA3⁺ tTregs, as this subset is severely reduced in the small intestine of IL2^−/− mice (Wohlfert et al. 2011). Furthermore, blocking IL2 in a mouse model of oral tolerance to ovalbumin impaired the conversion of adoptively transferred naive OT-II CD4⁺ T cells into pTreg cells in the mesenteric lymph nodes (MLNs) and Peyer’s Patches (PP) (Edwards et al. 2016). However, the role of IL2 in the induction of RORγt⁺ and RORγt⁻ pTregs has not been evaluated.

TGFβ gives an initial survival advantage to the Treg precursors in the thymus, though it fails to promote Foxp3 expression and is therefore not important for the induction of tTregs (Li et al. 2006; Marie et al. 2005, 2006; Ouyang et al. 2010; Tone et al. 2008; Konkel and Chen 2011). On the other hand, TGFβ has an unambiguously essential role in the differentiation of pTregs in the intestine, where it is present in copious amounts (Chen et al. 2003; Konkel and Chen 2011). TGFβ1-deficient mice develop lethal multi-organ lymphoproliferative disease, which mainly affects the gut. Notably, this pathology has considerable similarity with the one observed in the Foxp3-deficient mice (Kulkarni et al. 1993; Shull et al. 1992). TGFβ1-activated Smad3 binds to Foxp3 intronic enhancer CNS1 (Josefowicz et al. 2012; Tone et al. 2008; Xu et al. 2010). However, mice with an ablation of the Smad3-binding site in CNS1 did not result in reduction in the frequency of intestinal Tregs (Schlenner et al. 2012). Similarly, the frequency of colonic Tregs is not compromised in mice with the deletion of the cytokine receptor TGFβR1 specifically in Tregs (Konkel and Chen 2011). However, specific Treg subsets were not investigated in that study, and it is possible that while pTregs contracted in the absence of TGFβ signaling, the tTregs expanded to compensate. On the other hand, colitis in mice with TGF-βRII-deficient DCs is accompanied by reduced frequencies of mucosal and MLN Foxp3⁺ Tregs (Jamwal et al. 2020; Ramalingam et al. 2012).

TGFβ1 is secreted as an inactive precursor molecule combined with latency-associated peptide (LAP) and needs to be modified by integrins, such as αvβ8 and αvβ6 (Travis et al. 2007). Importantly, the expression of integrin β8 is highly upregulated in intestinal CD103⁺ DCs (Paidassi et al. 2011), and consequently CD103⁺ DCs deficient in αvβ8 have a compromised capacity for Treg induction in the MLNs (Worthington et al. 2011). Additionally, LAP could be cleaved by pro-protein convertase enzymes, such as Furin. Notably, Furin-deficient Tregs exhibit defective regulatory function in the naïve T-cell transfer-induced colitis (Pesu et al. 2008). Furthermore, in a mouse model of oral tolerance, a complex of GARP (glycoprotein A repetitions predominant) and latent TGFβ1 on Tregs has been found to be a major source of TGFβ1 needed for the induction of pTreg cells in the MLNs and PPs (Edwards et al. 2016). Overall, IL2 and TGFβ1 seem to be central in Treg homeostasis, so much, so that naïve T cells after in vitro stimulation in the presence of TGFβ and IL2 are induced to differentiate into inducible Tregs (iTregs) (Zheng et al. 2002, 2004).

Although the plasticity of Tregs under inflammatory pressure is well recognized, there are lingering controversies, mainly stemming from varying definitions, protocols, and starting cell populations. One aspect of Treg plasticity is related to reversible FoxP3 expression. The in vitro generated iTregs are generally recognized to have unstable expression of Foxp3 via a mechanism involving histone demethylation at the CNS2 element, resulting in quick reversion to Foxp3⁻ effector T cells (Kanamori et al. 2016). Another aspect of Treg plasticity is the described gain in IL17 production. IL6 is thought to contribute to the latter, although different Treg lineages seem to respond differently. nTregs can be induced to become Th17-like when exposed to IL6 in a TGFβ-dependent manner, but iTregs were resistant to this conversion under the same protocol (Zheng et al. 2008). Xu et al. also showed that CD4⁺CD25⁺FoxP3⁺ mouse splenocytes (constituting primarily nTregs) stimulated with IL6 turned into Th17 cells, even without the presence of exogenous TGFβ (Xu et al. 2007). Similarly, nTregs sorted from mouse thymus or spleens were converted to Th17 cells with CD3/IL6 stimulation, a conversion inhibited by atRA (Zhou et al. 2010). Human peripheral blood CD4⁺CD25^hiCD127^loFoxp3⁺ Tregs stimulated with high-salt medium lose their immunosuppressive function via gaining IFNγ expression, though only a very minor increase of IL17 expression was observed in this cell population. In a more recent mouse study, Luo et al. showed that high-salt diet in vivo and elevated NaCl concentration in vitro did not negatively affected pTregs or iTregs, respectively, but impaired the immunosuppressive function of tTregs and reduced FoxP3 expression (Luo et al. 2019). Thus, none of the Treg population should be considered as entirely stable or terminally differentiated. Their function and phenotype can be affected differently by various stimuli dependent on their lineage, tissue, and/or pathophysiological milieu.

Recently, IL33 has also been deemed instrumental in the induction of Tregs in the gut. IL33 is passively released by stromal or epithelial cells during cell necrosis or tissue damage, defining its role as an alarmin that alerts the immune system during trauma (Pichery et al. 2012). As mentioned earlier, intestinal GATA3⁺ tTregs, but not RORγt⁺ pTregs, have high expression of the IL33 receptor ST2. Likewise, in vivo administration of IL33 increases the number of both splenic Tregs and colonic RORγt⁻ Tregs, but not RORγt⁺ pTregs (Schiering et al. 2014; Sefik et al. 2015). Overall, IL33 seems to positively regulate tTregs but not pTregs in vivo. However, IL33 in combination with TGFβ1 has also been successfully able to induce the differentiation of iTregs (Schiering et al. 2014).

9.4.2. Host Factors: TNFRSF—NF-κB Axis

Survival at a barrier site such as intestine requires adaptation to the changing environmental milieu. Due to continuous exposure to microorganisms, Treg cells display an activated phenotype characterized by the core tumor necrosis factor receptor superfamily (TNFRSF) signature including expression of GITR, OX40, and TNFR2 (Miragaia et al. 2019; Vasanthakumar et al. 2017). The stimulation of TNFRSF members further activates the NF-κB signaling pathway and provides vital survival signals for Tregs (Miragaia et al. 2019; Salomon et al. 2018; Vasanthakumar et al. 2017). OX40 is important for survival, accumulation, and function of Treg cells in the colon (Griseri et al. 2010). Activation of RelA, a NFκB transcription factor, is associated with the adaptation of both tTregs and pTregs in the colon (Vasanthakumar et al. 2017). Overall, the TNFRSF—NF-κB axis may be a central signaling component in Treg maintenance in the gut.

9.4.3. Environmental Factors: Commensal Microbiota

The intestine, especially the colon, harbors abundant bacteria reaching 10¹² cells per gram of intestinal content. Bacteria remain in a complex relationship with the host immune system, and among their many functions, they profoundly affect mucosal T cells, including their differentiation into different subsets of effector or regulatory cells (Larmonier et al. 2015). As mentioned in the previous sections, the pTregs are induced, maintained, and “fine-tuned” directly by the commensal bacteria or by the metabolites produced by bacteria. While both RORγt⁺ and RORγt⁻ pTregs are affected by general presence or absence of microbiota, apparently, only the RORγt⁺ subset seems to be affected by qualitative changes in the colonic microbiome (Kim et al. 2016; Ohnmacht et al. 2015; Sefik et al. 2015; Ye et al. 2017). Some bacterial genera including Clostridium, Bacteroides, Bifidobacterium, Lactobacillus, and Helicobacter have been shown to promote pTregs in colon (Geuking et al. 2011; Atarashi et al. 2011, 2013; Chai et al. 2017; Kullberg et al. 2002; Sefik et al. 2015). Most of clostridia are gut commensals, with a few exceptions, such as Clostridium perfringens, Clostridium difficile, and Clostridium tetani, which are pathogenic and toxigenic. Oral administration of GF mice with 46 strains of commensal Clostridia showed strong promotion of differentiation, proliferation, and recruitment of Helios⁻RORγt⁺ Tregs in the colon (Atarashi et al. 2011). Similar experiments executed with 17 strains of Clostridia isolated from a healthy human adult also demonstrated a strong capacity to induce RORγt⁺Helios⁻ pTreg cells rather than GATA3⁺ tTregs in the colon of mice and rats (Atarashi et al. 2013). Clostridia-induced increase in frequency of colonic Tregs is also accompanied by an increase in the expression of IL10, CTLA4, and ICOS on pTregs (Atarashi et al. 2011, 2013) and increased intestinal production of sIgA which coats commensals and reduces T-cell activation by microbial antigens, such as flagellin (Cong et al. 2009). Faecalibacterium prausnitzii (cluster IV) is another member of the class of Clostridia and one of the most abundant residents in healthy humans (Godefroy et al. 2018; Sarrabayrouse et al. 2014). F. prausnitzii promotes the accumulation of Tr1 cells (IL10-producing CD4⁺CD8αα⁺ T cells) in the colon and the blood of healthy humans (Sarrabayrouse et al. 2014). Interestingly, relative abundance of F. prausnitzii is decreased in IBD patients (Machiels et al. 2014; Sokol et al. 2008, 2009). F. prausnitzii is capable of inducing IL10 secretion by healthy human DCs (Alameddine et al. 2019) which may also assist with reducing the overall inflammatory tone in the gut.

Apart from the Clostridia, another resident member of the human gut microbiome capable of impacting Treg homeostasis is Bacteroides fragilis. B. fragilis enhances the accumulation of IL10-producing Foxp3⁺ Tregs in the colon (Round et al. 2011). Monoassociations of mice with other Bacteroides spp., e.g. B. caccae, B. thetaiotaomicron, B. uniformis, B. finegoldii, or B. ovatus, have been shown to induce the accumulation of FOXP3⁺ cells, particularly Nrp1⁻RORγt⁺ pTreg cells in the mouse colon (Faith et al. 2014; Sefik et al. 2015). De novo generation of Tregs, with the capacity to suppress intestinal inflammation, was also observed in GF mice colonized with Bifidobacterium bifidum (Verma et al. 2018). Clostridia, B. fragilis, and B. bifidum have also been shown to increase the expression of CTLA4 on Tregs (Atarashi et al. 2011, 2013; Round et al. 2011; Verma et al. 2018). While it is not yet established if Clostridium and Bacteroides species function in a TCR-dependent manner, a pathobiont Helicobacter hepaticus has been shown to induce antigen-specific RORγt⁺ Tregs in the colon (Chai et al. 2017; Xu et al. 2018).

Microbes also stimulate the regulatory arm of the mucosal immune system via their components. For example, B. fragilis releases outer membrane vesicles (OMVs) containing polysaccharide A (PSA), the most abundant capsular polysaccharide expressed by B. fragilis. OMVs are taken up by cLP DCs and further elicit the conversion of CD4⁺ T cells into IL10-producing Foxp3⁺ Tregs via TLR2 (Round et al. 2011). Similarly, cell surface β-glucan/galactan (CSGG) polysaccharides from B. bifidum are important effector components able to induce Tregs via a DC-dependent and partially TLR2-dependent mechanism (Verma et al. 2018).

9.4.4. Environmental Factors: Dietary and Microbial Metabolites

The specific mechanism by which symbionts stimulate the accumulation of Tregs is not fully understood. It is believed that Clostridia spp. can function synergistically to stimulate IECs leading to Treg induction (Atarashi et al. 2013). One proposed mechanism of action is the cooperative production of short-chain fatty acids (SCFAs), such as butyrate, propionate, and acetate, through fermentation of dietary fiber by the commensals including Clostridium and Bacteroides spp. in the colon (Arpaia et al. 2013; Atarashi et al. 2013; Furusawa et al. 2013; Smith et al. 2013). Colonic SCFAs are absorbed by passive diffusion and by active H⁺-coupled transport via MCT1 (SLC16A1) and MCT4 (SLC16A3), two electroneutral monocarboxylate transporters, or by electrogenic Na⁺-coupled monocarboxylate transporters SMCT1 (SLC5A8) and SMCT2 (SLC5A12). Oral administration of butyrate, propionate, and acetate either individually or in combination led to an increase in the number of colonic Treg cells in SPF mice (Smith et al. 2013). Although in monoassociation studies, Sefik et al. could not find a correlation between luminal SCFA concentrations and the frequencies of RORγt⁺ pTregs (Sefik et al. 2015), this study did not take the combinatorial effects of SCFA producers into account. Upon reaching the colonic LP, SCFAs affect DCs and T cells directly using two mechanisms: via inhibition of histone deacetylases (HDAC) and via signaling through certain G protein–coupled receptors. Butyrate is known to participate in Treg differentiation by facilitating histone H3 acetylation in the promoter region and CNS1 and 3 of Foxp3 gene (Furusawa et al. 2013). Recognition of butyrate by G protein–coupled receptors such as GPR43and GPR15 expressed by colonic Treg cells and by GPR109A expressed by DCs and macrophages (Arpaia et al. 2013; Atarashi et al. 2013; Furusawa et al. 2013; Singh et al. 2014; Smith et al. 2013) also promotes Treg differentiation. Interestingly, SCFA treatment increased Helios⁺ Treg numbers in GF mice, indicating that SCFAs also promote the expansion of the tTregs present in the colonic LP (Smith et al. 2013).

Aryl hydrocarbon receptor (AhR) is a nuclear sensor expressed by Tregs that helps them sense and react to compounds that act as AhR ligands. Intestinal pTregs have higher expression of AhR than Tregs in any other tissue (Ye et al. 2017). Dietary tryptophan, an essential amino acid, is metabolized by IDO (indoleamine 2,3-dioxygenase) and TDO (tryptophan 2,3-dioxygenase) expressed in IECs and DCs, into several AhR ligands including kynurenine. Kynurenine enhances Foxp3⁺ Treg generation in vitro in the presence of TGFβ1 (Mezrich et al. 2010). Gut microbes like Lactobacillus spp. can also metabolize tryptophan into many AhR ligands (Gao et al. 2018). AhR is not only important for the induction of Tregs, but it also contributes to their regulatory function. Ahr-expressing Tregs showed enhanced in vivo suppressive activity compared with Ahr-deficient Tregs in a T-cell transfer model of colitis (Ye et al. 2017).

Another dietary metabolite that drives the Treg development in intestine is the all-trans retinoic acid (atRA), a bioactive form of vitamin A. Dietary vitamin A, specifically retinol, is absorbed by the intestinal epithelial cells via passive diffusion. Although epithelial cells are capable of synthesizing atRA from dietary vitamin A (D’Ambrosio et al. 2011), a process regulated by commensal bacteria (Grizotte-Lake et al. 2018), the vast majority of studies to date focused on dendritic cells, macrophages, and eosinophils as sources of atRA for iTreg induction. Aldehyde dehydrogenases ALDH1A1, ALDH1A2, and ALDH1A3 are the rate limiting enzymes in the synthesis of atRA (Bazewicz et al. 2019). Of those three, CD103⁺DCs express primarily the ALDH1A2 isoform (Coombes et al. 2007). CD103⁺ DCs are chiefly equipped to promote Tregs in the gut by the activation of latent TGFβ1 via integrin αVβ8. Thus, CD103⁺ DC-derived RA, along with TGFβ, converts naive FoxP3⁻CD4⁺ T cells into FoxP3⁺ Tregs in both human and mouse (Coombes et al. 2007; Kang et al. 2007; Mucida et al. 2007; Sun et al. 2007). These Tregs also display an enhanced upregulation of the mucosal tissue-homing receptors C-C chemokine receptor type 9 (CCR9) and integrin β7 along with being potent suppressors of effector T cells in a colitis model (Coombes et al. 2007; Kang et al. 2007; Mucida et al. 2007; Sun et al. 2007). These RA-induced human FoxP3⁺ cells were also chemotactically responsive to CCR9 ligand CCL25, a chemokine specifically expressed by intestinal epithelial cells (Kang et al. 2007). Furthermore, CNS1 non-coding element in the FoxP3 gene, which is required for the generation of pTregs, contains binding sites for the heterodimer of retinoic acid receptor (RAR) and retinoid X receptor (RXR). Upon binding of RA to RAR and RXR, CNS1 undergoes histone acetylation leading to the prolonged and stable expression of FoxP3 in Tregs (Xu et al. 2010). atRA was also shown to promote iTreg development and maintenance via an ERK1/2-dependent mechanism and increased histone methylation and acetylation within the CNS2 element of Foxp3 gene locus (Lu et al. 2011).

Additionally, it is also reported that RA is able to suppress the IL6-induced conversion of Foxp3⁺ Tregs into inflammatory Th17 cells, a mechanism likely mediated by the retinoic acid receptor α (Elias et al. 2008). Based on experiments with mice fed with a vitamin A–deficient diet or treated with an inhibitor of the RA receptor, vitamin A seems to affect the development of RORγt⁺ pTregs cells and not Helios⁺ tTreg cells in the gut (Ohnmacht et al. 2015). While RORγt⁺ Tregs were restored upon feeding RA, it did not “rescue” the defective generation of RORγt⁻ Tregs (Kim et al. 2016). Notably, an RA-reactive human Treg cell subset expressing natural killer cell receptor CD161 and RORγt has been described recently (Povoleri et al. 2018).

While vitamin A plays the most prominent role in the development of Tregs, there are other vitamins that have been reported to affect this lymphocyte population. Dietary vitamin D3 is metabolized to 1,25-dihyroxyvitamin D3 which binds to the nuclear vitamin D receptor element (VDRE) within the +1714 to +2554 nt non-coding CNS region of the human Foxp gene (Kang et al. 2012). FR4, the receptor for vitamin B9 (folic acid), is highly expressed by Tregs (Yamaguchi et al. 2007) and is known to promote survival of colonic Tregs by upregulating the anti-apoptotic factor BCL2 (Kinoshita et al. 2012). Consistent with this observation, mice fed with a vitamin B9–deficient diet displayed a marked reduction in colonic pTregs and a higher susceptibility to intestinal inflammation (Kinoshita et al. 2012).

9.5. Suppressive Function Mediated by Treg in the Gut

Tregs can exert suppressive function towards a variety of immune cells, such as NK, CD8⁺, and CD4⁺ T lymphocytes and antigen-presenting cells (APC). Inhibition of T conventional cells can occur indirectly, i.e., when inhibition involves influencing the activation status of APC, or directly, i.e., in absence of APC. Tregs can suppress intestinal immune responses to maintain homeostasis and exert their functions through various mechanisms, mostly determined on the basis of in vitro assays. The best-known suppressive mechanisms to be discussed here are (1) cytokine-mediated inhibition, (2) expression of inhibitory surface molecules, and (3) cytolysis of target cells. Species-specific mechanisms may complicate data interpretation, and findings identified in murine models may not correspond to those found in human Tregs.

9.5.1. Secretion of Inhibitory Cytokines

FoxP3⁺ Tregs and Tr1 cells produce high levels of IL10, which can target several cell types in the gut, among them the effector immune cells, to inhibit their expansion and to promote immunosuppression. IL10 also acts in an autocrine manner as a proliferative factor for Tregs, although its requirement for FoxP3⁺ Tregs and Tr1 maintenance in vivo is unlikely, since both Treg subsets develop normally in the absence of IL10 or IL10R (Maynard et al. 2007; Chaudhry et al. 2011). Even though IL10 can be secreted by many cells of the innate and adaptive immune systems, the importance of IL10 originating in Tregs for the gut homeostasis is illustrated by the development of exacerbated intestinal inflammation when FoxP3⁺ Tregs lacked the ability to produce IL10, likely due to their inability to suppress pathogenic Th17 cell responses (Asseman et al. 1999; Chaudhry et al. 2011). The high concentration of IL10 produced by mucosal Tr1 enables the suppression of IL1β and TNFα released from human myeloid cells (Cook et al. 2019). In the gut homeostasis, IL10 gene has been identified as susceptibility locus in ulcerative colitis (UC) (Franke et al. 2008) and mutations on IL10RA and IL10RB receptors result in severe and early onset IBD (Glocker et al. 2009; Moran et al. 2013). In the murine colon, RORγ⁺ FoxP3⁺ Tregs have been reported to be the main producers of IL10 (Ohnmacht et al. 2015).

The involvement of Treg-derived TGFβ in immune tolerance, inhibition of acute inflammation, and promotion of wound-healing process renders it a pivotal cytokine in the immune regulation in the gut. Soluble and membrane bound TGFβ are produced in high amounts by Tr1 and FoxP3⁺ Tregs in the gut (Powrie et al. 1996; Levings et al. 2002). TGFβ also contributes to the differentiation of Tregs, converting naïve CD4⁺ T cells into FoxP3⁺ Tregs (Chen et al. 2003; Zheng et al. 2002), but other mediators are required for Tregs to fully differentiate since the induction of FoxP3 expression by TGFβ alone does not necessarily confer immunosuppressive abilities to Tregs (Tran et al. 2007). Production of TGFβ by Treg is, however, necessary to prevent intestinal inflammation (Powrie et al. 1996) as much as expression of TGFbRII by the T conventional cells (Gorelik and Flavell 2000), indicating the essential role of Treg-derived TGFβ in controlling pathogenic T-cell response intestinal inflammation.

Tr1 expanded in vitro, but not FoxP3⁺ Tregs, can also secrete IL22 to promote intestinal barrier function (Cook et al. 2019). Even though IL35 has been described as an important immunoregulatory cytokine secreted by murine Tregs (Collison et al. 2007), human FoxP3⁺ Tregs in resting or activated state appear not to express IL35 (Bardel et al. 2008).

9.5.2. Expression of Cell Surface Inhibitors of Inflammation

Human Tregs express cell surface molecules that enable the suppression of immune function in target cells via a contact-dependent (e.g., CTLA4, LAG3, and PD1) or contact-independent manner (e.g., IL2R and CD39). The contact-dependent suppression involves binding of the inhibitory receptors expressed on the Treg surface to stimulatory and co-stimulatory receptors expressed on the surface of target cells, ultimately suppressing immunostimulatory intracellular cascades in target cells. The successful antigen presentation by APC requires the antigen-TCR binding in the context of major histocompatibility complex II (MHCII), in coordination with the triggering of CD28 receptors on T cell by co-stimulatory receptors from the B7 family on APC. B7.1 (CD80) and B7.2 (CD86) can bind to a variety of molecule members of the CD28 family: CD28 and ICOS act as positive regulators of T-cell function while CTLA4 and PD1 act as inhibitors. B7.1 and B7.2 are also expressed on T_eff cells and can be targeted for inhibition through contact-dependent mechanisms.

9.5.2.1. Contact-Dependent Inhibition

Cytotoxic T-lymphocyte-associated protein 4 (CTLA4, CD152) is highly expressed on FoxP3⁺ Treg and Tr1 cells relative to the conventional CD4⁺ T cells. It is a key regulatory molecule, since null or Treg-specific knockout of CTLA4 result in autoimmune disease (Takahashi et al. 2000; Wing et al. 2008). In vitro suppression of T_eff cell proliferation by Tregs (in presence of APC) was shown to be dependent on CTLA4 in murine and human cells (Tang et al. 2004; Birebent et al. 2004) and was mediated by CD80/CD86 downregulation on APC (Cederbom et al. 2000). However, production of inflammatory cytokines does not seem to be affected by Tregs’ CTLA4. Expression of lymphocyte activation gene 3 (LAG3) in human Tregs also contributes to their suppressive function (Wang et al. 2018). LAG3 is a CD4 homolog expressed on FoxP3⁺ and Tr1 T cells and interacts directly with MHCII complex acting as a decoy receptor to inhibit its interaction with the TCR, downregulate cell proliferation and secretion of inflammatory cytokines IFNγ, IL2, and TNFα by APC and B cells (Hannier et al. 1998). Deficiency in LAG3 renders Tregs inefficient to inhibit T-cell proliferation and unable to prevent chronic intestinal inflammation (Huang et al. 2004). PD1 (CD279) is also expressed on both Treg subsets and binds to PD-L1 (B7-H1), PD-L2 (B7-DC), and B7.1. PD1 signaling inhibits TCR-signaling-induced cell cycle progression and TCR expression on CD8⁺ T cells (Karwacz et al. 2011), and blockade of the PD1/PD-L1 pathway can abrogate Treg-mediated immunoregulation (Kitazawa et al. 2007).

9.5.2.2. Contact-Independent Inhibition

In humans, suppressive Tregs express high levels of CD25. CD25 is a subunit of the high-affinity heterotrimeric IL2 receptor and is also transiently expressed on activated T cells. CD25 expression is a part of the suppressive mechanism of Tregs, supposedly through the scavenging of IL2 and reduction of its effects on T cells (Barthlott et al. 2005). Though the proximity between cells is necessary for inhibition mediated by CD25, contact is not required. Exogenous IL2 abrogates Treg-mediated suppression of T_eff in human and mice (Oberle et al. 2007; Pandiyan et al. 2007). Recent reports link the IL2 scavenging by Tregs to induction of apoptosis on T_eff cells (Pandiyan et al. 2007), but further studies are required to prove the existence of such regulatory mechanism in humans.

Expression of CD39 on the surface of FoxP3⁺ Tregs also provides an important immunoregulatory mechanism. Catalysis of ATP to AMP by CD39 (ecto-nucleoside triphosphate diphosphohydrolase 1, E-NTPDase1), and its subsequent hydrolysis into adenosine by CD73 (ecto-5′-nucleotidase, NT5E) regulates the extracellular levels of ATP and adenosine and their effects on immune cells (Su et al. 2019). Since human Tregs do not express high levels of CD73 (in contrast to their murine counterparts), the AMP conversion to adenosine in humans is thought to result from the expression of CD73 on neighboring cells (Dwyer et al. 2010). ATP binding to purinergic receptors such as P2X7 and P2Y2 expressed on the surface of immune cells induces the secretion of inflammatory cytokines and a DC-mediated priming of intestinal T_eff cells toward the Th17 phenotype (Pandolfi et al. 2016; Atarashi et al. 2008). On the other hand, activation of A2A receptors by adenosine induces immune anergy and promotes intestinal homeostasis (Ye and Rajendran 2009). Fifty percent of the human Tregs express CD39 (Borsellino et al. 2007), and this population is highly suppressive through the generation of adenosine (Mandapathil et al. 2010). CD39 polymorphisms in humans are associated with IBD susceptibility (Friedman et al. 2009). ATP/adenosine balance by CD39⁺ Tregs is of relevance to the intestinal inflammation, where high concentrations of ATP produced by the microbiota reach the LP immune cells as a consequence of intestinal epithelial barrier injury (Atarashi et al. 2008).

Inducible co-stimulator (ICOS) has a dual and balancing role to sustain T-cell activation and effector functions and to promote Treg differentiation and suppression. ICOS expression on human Tregs is potentially involved in contact-dependent and -independent immunosuppression in peripheral tissues. High expression of ICOS on natural FoxP3⁺ Tregs present in thymus, PBMC, and peripheral tissues endows Tregs with the capacity to produce high levels of IL10 and TGFβ to suppress DC and T-cell function, respectively, whereas ICOS⁻FoxP3⁺ Tregs play a suppressive function mainly through the secretion of TGFβ1 (Ito et al. 2008). ICOS contact-dependent CD4⁺ T suppression is mediated by the upregulation of CTLA4 expression on human Tregs (Zheng et al. 2013). Some of the findings in the literature support the role of the microenvironment in the differentiation of ICOS⁺ Tregs. These cells are the main suppressive subset in human melanoma (Strauss et al. 2008). In the murine gut, colonic LP ICOS⁺ Tregs have a strong immunosuppressive profile (Nakanishi et al. 2018) and are stimulated locally by human microbiota (Atarashi et al. 2013).

9.5.3. Cytolysis

A less studied mechanism of Treg-mediated immune suppression in human is the secretion of cytolytic enzymes and killing of target cells. Human CD25⁺ Tregs generated in vitro from CD4⁺ T cells stimulated with antibodies to CD3/CD46 and with IL2, express granzyme B and induce cell death in autologous cells in a perforin/granzyme-dependent manner (Grossman et al. 2004a). Although CD25^high Tregs from human peripheral blood activated in vitro did not express granzyme B, they expressed granzyme A and presented a cytolytic activity toward various autologous immune cells in a perforin-dependent but FasL-independent fashion (Grossman et al. 2004b). The target cells tested included activated CD4⁺ and CD8⁺ T cells, CD14⁺ monocytes, and dendritic cells. However, iTreg-mediated suppression of B cells seems to be independent of cytolysis (Xu et al. 2016). Two studies reported the presence of FoxP3⁺ Tregs expressing granzyme B in the colon of patients with colorectal cancer (Sun et al. 2020) and in the lungs of RSV-infected mice (Loebbermann et al. 2012). Very few to none of these cells could be found in lymphoid organs or circulating Tregs, suggesting that granzyme B expression is acquired locally by Tregs. These findings, summarized in Fig. 9.2, indicate the participation of perforin/granzyme-dependent cytotoxicity in Treg-mediated suppression in the mucosal surfaces.

Fig. 9.2.

Mechanisms of immunosuppression in human regulatory T cells. Tregs suppress inflammatory cells through secretion of anti-inflammatory cytokines, expression of inhibitory surface molecules and cytolysis of target cells. Secretion of IL22, IL10 and TGFβ, as well as cytolytic Granzyme A/B and Perforin leads to suppression of effector functions of a variety of immune cells. ATP hydrolysis by CD39 also limits the level of such inflammatory signaling molecule in extracellular medium. Expression of LAG3 and CTLA4 constitute negative regulators of APC activation, through MHC competition and inhibition of CD80/86 signaling respectively. High expression of IL2Rα scavenges the extracellular IL2 available for T_eff cells, while PD-1 expression inhibits TCR expression and signaling in effector T cells. ICOS induces secretion of TGFβ1 and CTLA4 expression

9.6. Treg During Intestinal Diseases

Tregs are central regulators of the intestinal immune response and inflammation in the gut. Through the described mechanisms, Tregs suppress immune activation and control the intensity of inflammatory responses in the gut. A breakdown of these tolerogenic mechanisms is usually observed under pathological conditions involving exacerbated inflammation in the intestinal mucosa, e.g., in inflammatory bowel disease, celiac disease, or necrotizing enterocolitis (NEC). On the other hand, the function of Tregs can be detrimental under certain circumstances in which the immune response is necessary to eliminate infectious agents or oncogenically transformed cells. In this topic, we will focus on intestinal diseases where a dysfunction of regulatory T cells has been implicated.

9.6.1. Inflammatory Bowel Disease

IBD is a group of chronic inflammatory diseases resulting from exacerbated immune response in the gut mucosa. There is a consensus that the inappropriate immune response driven against commensal intestinal microbes is the key to IBD etiology and is affected by both the environmental factors and genetic susceptibility of the host. Ulcerative colitis (UC) and Crohn’s disease (CD) are the two major types of IBD, distinguished on the basis of histological and endoscopic findings, distribution, and clinical complications (Molodecky et al. 2012). Several lines of evidence support the role of T cells in IBD; however, the adaptive immune profile during CD and UC differs. CD is noted for the involvement of Th1-related cytokines TNFα, IFNγ, and IL12, whereas UC is driven by Th2 and Th9 cytokines such as IL5, IL13, and IL9 (Muzes et al. 2012). Th17-related cytokines, IL17A, and IL23 are increased in intestinal tissue in both UC and CD (Nemeth et al. 2017) and contribute to the disease pathogenesis in these patients.

Multiple human genome-wide association studies (GWAS) have identified a remarkable proportion of single-nucleotide polymorphisms (SNPs) in genes that affect T-cell function in patients with IBD. Among them, polymorphisms in genes linked to the IL10 signaling pathway are associated with severe early onset of IBD (Farh et al. 2015; Glocker et al. 2009). IL10 signaling is essential to maintain homeostasis and mucosal barrier function and to the control of effector and suppressor cells. As described earlier in this chapter, Tr1 and FoxP3⁺ Tregs are the two major producers of IL10 in the gut. The pivotal role of FoxP3⁺ Tregs in maintaining intestinal homeostasis is exemplified by the presence of severe intestinal inflammation in patients with FoxP3 mutations, who suffer with global failure in Treg development (Bennett et al. 2001). Interestingly, patients with IBD do not have a quantitative deficiency in frequency of FoxP3⁺ Tregs in the gut. On the contrary, an increase in this population has been observed in the mucosa and regional lymph nodes of CD and UC patients (Reikvam et al. 2011; Yu et al. 2007). Ex vivo, CD4⁺CD25⁺ cells isolated from UC patients showed suppressive activity, and the authors suggested that this activity may either be insufficient or impaired in vivo, in the face of overwhelming mucosal inflammation (Yu et al. 2007).

Helios+ and Helios− FoxP3⁺ cells in the mucosa of IBD patients were found to express similar levels of immunoinhibitory receptors CD25, CD39, CTLA4, TIGIT, and PD-1 compared to control patients (Lord et al. 2015). Analyses of circulating Tregs in IBD can be controversial, and the frequency of FoxP3⁺ Tregs and Tr1 cells has been found either increased (Vitale et al. 2019) or decreased (Khalili et al. 2018). In murine adoptive naïve T-cell transfer model of colitis, a reliable and representative model of IBD (te Velde et al. 2007), the administration of a T-cell population containing IL10-producing Tregs, protected from intestinal inflammation (Asseman et al. 1999). CTLA4 was found to be mutated in mucosal immune cells in early onset CD (Zeissig et al. 2015). Impaired dimerization and binding of the mutated CTLA4 to CD80 could explain the loss of Treg regulatory mechanisms in IBD. Chao et al. (2018) showed that lack of CTLA4 in CD4⁺ T cells in mice impairs differentiation of follicular Tregs and triggers the accumulation of autoantibodies in intestinal epithelial cells. This defect exacerbated the immune response to MCMV (mouse cytomegalovirus), resulting in acute intestinal injury and death.

Another mechanism driving impaired immunosuppression in IBD may be related to FoxP3⁺ Treg survival in the inflamed intestine as a consequence of the hyperinflammatory environment. It has been reported that Tregs from the inflamed tissue of IBD patients show evidence of enhanced apoptosis, which can be reversed by anti-TFNα treatment (Veltkamp et al. 2011). In line with the evidence of altered Treg survival during IBD, polymorphisms in genes encoding the IL2RA and IL2RB subunits, indispensable for Treg survival in peripheral tissues, were identified in patients with CD and UC (Bouzid et al. 2013). ST2/IL33 axis has also been shown to enhance Treg stability and/or survival. Under homeostasis, IL33 binds to suppressor of tumorigenicity 2 (ST2) highly expressed on colonic Tregs, stimulates FoxP3 and GATA3 expression, and promotes Treg function through enhancing TGFβ1-mediated differentiation, also an inducer of survival pathway in T cells (Schiering et al. 2014). IL33 signaling in DCs may indirectly contribute to this process, by stimulating DC production of IL2 (Matta et al. 2014). Increased levels of sST2, an extracellular ST2 decoy receptor for IL33, in the mucosa and serum of IBD patients is positively associated with the disease severity, suggesting that this axis maybe be disrupted during chronic inflammation in patients with IBD (Boga et al. 2016). In sepsis-surviving patient, IL33 contributes to long-term immunosuppression by expanding the Tregs (Nascimento et al. 2017). However, in UC patients, increased IL33 mRNA expression was reversed in remission induced with infliximab, a TNFα inhibitor, which could be interpreted that IL33 may have a dual role in IBD patients (Gundersen et al. 2016; Chen et al. 2020; Pastorelli et al. 2013).

In addition to a reduced survival observed in Tregs from patients with IBD, alterations in Treg function and phenotype have been reported. Zhou et al. showed that FoxP3⁺ Tregs have reduced FoxP3 expression and acquire inflammatory phenotype when exposed to an inflammatory milieu (Zhou et al. 2009). Loss of FoxP3 interaction with the epigenetic enzyme Enhancer of Zeste Homolog 2 (EZH2) observed in mucosal Tregs of IBD patients may result from Treg exposure to IL6 (Bamidele et al. 2019) and may lead to an impaired repression of inflammatory genes, normally targeted by FoxP3/EZH2 complex in CD4⁺ T cells (Sarmento et al. 2017). Sanctuary et al. (2019) described a mechanism for TNFα-induced loss of Treg immunosuppressive function. According to the authors, TNFα induces miR-106a microRNA in both humans and mice, which through NFκB promoter binding suppresses post-transcriptional regulation of IL10 release by Tregs.

In homeostatic conditions, intestinal FoxP3⁺ Tregs express RORγT. While this population is highly immunosuppressive, it may also display considerable plasticity and gain the ability to produce IL17A in inflammation associated with CD and UC. IL17A producing Tregs from IBD patients showed defective suppressive function, and differences in cytokine requirements for that transition were observed in patient peripheral blood cells from CD and UC (Ueno et al. 2013). In CD patients, IL1β/TGFβ/IL23 are more effective in inducing IL17A⁺ FoxP3⁺ Tregs, while cells from UC patients required IL21 and TGFβ (Ueno et al. 2013). Patients with IBD also demonstrated increased Treg/Th1 crossover populations in T-bet⁺FoxP3⁺ Treg (Li et al. 2017), but the pathways involved in acquisition of T-bet in human intestinal FoxP3⁺ Tregs and their role during IBD have not been investigated. To complicate matters, another study showed that despite producing IL17, IL17⁺CD161⁺ Tregs enriched in the CD mucosa remained highly suppressive and were enriched for wound healing genes (e.g., PDGFA and CFS2), including soluble mediators (e.g., IL17A and IL22) known for accelerating epithelial barrier healing (Povoleri et al. 2018).

The mucosal inflammatory milieu of IBD may also impair the responsiveness of the effector cells to Treg-mediated suppression. Inflammatory cytokines and inducers of NFκb in T lymphocytes, such as TNFα, have anti-apoptotic effects in these cells (Dudley et al. 1999) and could lead to pathogenic T-cell resistance to mechanisms of cell death induced by Tregs during IBD. Such resistance could promote the persistence and perpetuation of pathogenic CD4⁺ T cells in the gut tissue. Mechanisms of resistance to TGFβ have also been identified in the LP effector cells during IBD. High levels of Smad7 produced by mucosal T lymphocytes in patients with IBD, likely as a consequence of increased TNFα, IL1β, and IFNγ expression (Ulloa et al. 1999), turn pathogenic T_eff cells insensitive to regulation by TGFβ (Laudisi et al. 2016), despite the increased levels of TGFβ in IBD tissues (Babyatsky et al. 1996).

Most of the knowledge on Treg dysfunction in IBD comes from studies on FoxP3⁺ Treg subset, and the characteristics and function of Tr1 in patients with IBD have yet to be further investigated. A recent report has identified IFNγ⁺ IL10-producing Tr1 as CCR5⁺PD-1⁺ cells in the human intestine but did not indicate altered frequencies of these population in CD or UC. However, Tr1 showed reduced IL10 expression when isolated from the IBD mucosa, a decrease that could be reproduced by treatment with of IL-23 and IL1β (Alfen et al. 2018).

9.6.2. Celiac Disease

Celiac disease is a T-cell-mediated disease triggered by ingestion of gliadin, the main component of the gluten, in genetically predisposed individuals. Gliadin-derived peptides activate T helper (Th) cells to produce Th1 and Th17 signature cytokines IFNγ (Nilsen et al. 1998) and IL17A (Fernandez et al. 2011) that contribute to the expansion and maintenance of the inflammatory response in the gut of individuals with celiac disease. A break of tolerance to dietary gliadin, and the known role of Tregs in immune tolerance to dietary components, prompted several groups to investigate the Treg population patients with celiac disease. The findings offered some parallels and similarities with IBD. Like in IBD, the numbers of FoxP3⁺ Tregs are increased in the intestinal mucosa of celiac patients with active disease (Cianci et al. 2012), and its local expansion after gliadin stimulation (Zanzi et al. 2011) suggested that it is an insufficient attempt at reigning in the T_eff cells during this disease. Celiac intestinal mucosa also harbors Tr1 cells, able to produce IL10 and TGFβ and to suppress proliferation of pathogenic T cells in vitro (Gianfrani et al. 2006). In fact, upregulation of these two immunosuppressive cytokines by Tregs was observed along with the exacerbated proinflammatory immune response found in celiac disease (Lahat et al. 1999). In vitro studies support the regulatory role of IL10 in the mucosal T cells from celiac patients by inhibiting IFNγ production by gliadin-specific T_eff cells (Salvati et al. 2005).

The negative regulation of the immune response exerted by Tregs is not enough to control the ongoing immune response in celiac patients. Circulating FoxP3⁺ Tregs from children with celiac disease show high expression of CD101 and CD129 (Akesson et al. 2015) and 80% of the circulating FoxP3⁺ Tregs from celiac patients express high levels of CD39 (Cook et al. 2017), all indicators of their suppressor activity. FoxP3⁺ Tregs from the intestine of celiac patients showed no impaired function in vitro (Zanzi et al. 2011), but Cook et al. showed that circulating gluten-specific FoxP3⁺CD39⁺ Tregs from celiac disease patients can have reduced suppressive function in response to a polyclonal stimulus (Cook et al. 2017). These results suggest that the intestinal and the inflammatory environment established during the course of the disease may locally impair crucial regulatory mechanisms afforded by Tregs. The overexpression of intestinal IL15 in celiac disease could contribute to this phenomenon. IL15 impairs TGFβ signaling in biopsies from active celiac disease (Benahmed et al. 2007), dampens Treg suppressive activity (Zanzi et al. 2011), and makes T effector cells refractory to the regulatory effects of Tregs through the activation of phosphatidylinositol 3-kinase (PI3K) pathway (Ben Ahmed et al. 2009). Additionally, microbiome-derived metabolites have shown to modulate alternative splicing and the expression of FoxP3 isoforms in intestinal biopsies from patients with active celiac disease, where increased expression of FoxP3 shorter D2 isoform over full length has been observed (Serena et al. 2017). The alternatively spliced isoform FoxP3 D2, thought to be induced by butyrate produced by bacteria expanding during the active celiac disease in presence of IFNγ, is less effective in downregulating the Th17 differentiation. In this scenario, the switch to FoxP3 isoform D2 in patients with celiac disease could promote local Th17 immune response (Serena et al. 2017). Therefore, the microenvironment in the intestinal tissue of patients with celiac disease might orchestrate the differentiation and/or functionality of Tregs, thus restricting their activity.

9.6.3. Necrotizing Enterocolitis

Necrotizing enterocolitis (NEC) is an immunopathology that most commonly affects infants born prematurely. The disease is characterized by dysregulated inflammatory host responses to luminal bacteria with increased levels of inflammatory mediators, such as TNFα and plateletactivating factor (PAF) in the small intestinal tissue and serum (Caplan et al. 1990). The inflammatory injury in NEC causes a variable spectrum of damage to the intestinal tract, from mucosal injury to full-thickness necrosis. Multiple factors are involved in the precipitation of NEC, including formula feeding, ischemic/hypoxic injury, and abnormal bacterial colonization.

The population of FoxP3⁺ Tregs develop early during gestation. They are detected as early as 23 weeks of gestational age in large and small bowel. Treg numbers are stable during the later phases of development and do not change with increasing intrauterine development or postnatal age (Weitkamp et al. 2009). The immaturity of the immune system in preterm babies is thought to be a critical pathogenic factor in NEC. Several studies have found important correlations between preterm birth and growth restriction with Treg function in the intestinal tissue. The FoxP3⁺ Treg frequency in the circulating blood of preterm infants inversely correlated with gestational age at birth, indicating a contraction in this population after birth in preterm children (Qazi et al. 2020). Treg suppressive index, which reflects the expression level of FoxP3, is also decreased in preterm infants (Mukhopadhyay et al. 2014). Their homing markers may also be affected in preterm children (Qazi et al. 2020) and children with NEC (Weitkamp et al. 2013). In infants with growth restrictions, Mukhopadhyay et al. found a markedly reduced proportion of circulating Tregs in the cord blood (Mukhopadhyay et al. 2014). In this vulnerable group of infants, circulating FoxP3⁺ Tregs also showed a reduced expression of the gut homing integrin α4β7, which could also be an indicator of reduced capacity of Tregs to reach the intestinal mucosa (Qazi et al. 2020). Therefore, lower frequency, impaired homing, and/or impaired function of Tregs observed in preterm children may be implicated, or at least contribute to the pathogenesis of NEC.

In NEC, the inflammatory microenvironment may contribute to the impairment of Treg differentiation and expansion, which could perpetuate the inflammation. Flow cytometry analysis has shown lower frequency of FoxP3⁺ Treg cells in ileum (Weitkamp et al. 2013) and peripheral blood mononuclear cells of infants with NEC (Pang et al. 2018a). The frequency of intestinal FoxP3⁺ Tregs was recovered in tissue from healed postoperative NEC (Weitkamp et al. 2013, suggesting that Tregs from children with NEC have no intrinsic failure in expansion and accumulation in intestinal tissue. Studies of Tregs in infants with NEC need to be interpreted with caution due the diversity of markers used to identify these cells in different investigations. Detecting FoxP3 expression in Tregs is a useful tool to characterize the frequency of Tregs in the inflamed tissue, but it requires cell fixation and does not permit downstream functional analysis. Because of that, isolation of human Tregs for in vitro assays is usually based on high expression of IL2Rα (CD25). Human CD4⁺CD25^+/hi Tregs from NEC infants showed reduced expression of CTLA4, LAG3, and Helios, reduced capability to suppress CD4 T_eff proliferation in vitro, and remarkably failed to suppress IL17A expression in these cells (Pang et al. 2018a). An imbalance caused by an increased ratio of Th17/Treg cells in the LP was shown to contribute to excessive inflammatory response in NEC (Egan et al. 2016), and the restoration of this balance after Treg induction by retinoic acid administration decreased NEC severity (Nino et al. 2017). IL6, which inhibits TGFβ-induced Treg differentiation and favors Th17 lineage, was found to be increased in the small intestine affected by NEC and could participate in the contraction of Tregs and in perpetuating the disease (Weitkamp et al. 2013). Indeed, IL6 neutralization restored FoxP3⁺ Tregs and ameliorated experimental NEC in mice (Ma et al. 2019). Monocytes in NEC patients were suggested to play a role in Th17/Treg imbalance by promoting the differentiation of RORγt-expressing Th17 cells via production of IL6 (Pang et al. 2018b). The plasticity of Treg phenotype in hyperinflammatory environment, as we discussed earlier in the chapter, could also contribute to Treg dysfunction in NEC patients. IL17⁺ Tregs were increased in peripheral blood of NEC patients, and this population could be promoted by in vitro Treg exposure to IL6 (Ma et al. 2019).

9.6.4. Graft-Versus-Host Disease (GVHD)

GVHD is a syndrome commonly associated with allogeneic hematopoietic stem cell transplantation (HCT), caused by the recognition of the host as foreign by the immune cells that remain in the donated tissue (graft). The naïve donor T cells migrate to target organs such as the colon, small bowel, liver, and skin where they cause tissue damage (Wysocki et al. 2005). These tissues are potentially attractive sites for the donor T cells due the release of danger-associated molecular patterns (DAMPs), the presence and epithelial translocation of pathogen-associated molecular patterns (PAMPs), and inflammatory cytokines induced during irradiation and/or chemotherapy. Host’s activated APC activate donor T cells, leading to their proliferation and differentiation toward Th1 and Th17 cells, which cause inflammation and damage at mucosal tissues, including the gut. In human allogeneic transplant, localization of Foxp3⁺ Tregs in the gut has shown negative correlation with the severity of intestinal inflammation in GVHD (Rieger et al. 2006). Increased frequencies of Tregs with gut-homing phenotype (α4β7⁺) is associated with reduced gut GVHD in patients subjected to neutrophil engraftment (Engelhardt et al. 2011, 2012), indicating that Tregs inhibit immune responses at mucosal sites during GVHD. However, the involvement of Tregs in homeostasis during this syndrome may differ at different sites of the gastrointestinal tract, since no correlation between Treg frequency and the histological or clinical severity of gastrointestinal GVHD was observed in gastric antral biopsies in allogeneic hematopoietic stem cell transplantation (allo-HCT)recipients (Lord et al. 2011). Both tTreg and pTreg cells are important to mediate immunosuppression in the intestine during GVHD, and their transfer prior or along with the donor T cells has shown to diminish the incidence of GVHD and improved survival (Edinger et al. 2003). The literature about the use of Treg induction or infusion as means of GVHD prevention will be discussed in the next section of this chapter.

9.6.5. Tregs in Colorectal Cancer

Cancers are not just masses of transformed cells, but are complex aggregates with non-transformed cells, cytokines, chemokines, growth factors, and inflammatory and matrix remodeling enzymes. Colorectal cancer (CRC) remains one of the leading causes of death worldwide. Besides genetic predisposition and environmental factors, IBD-associated inflammation has been linked to CRC in an “inflammation-dysplasia-carcinoma sequence” (Zisman and Rubin 2008). The tumor microenvironment is thus a result of complex host-tumor interactions on multiple molecular and cellular levels. Immune cells contribute to this microenvironment with the primary goal to efficiently recognize and destroy neoplastic cells in a process known as tumor-immunosurveillance (Dunn et al. 2004). Tumor infiltrating lymphocytes (TILs) like CD8⁺ cytotoxic T lymphocytes (CTLs) are the major key players in tumor-immunosurveillance. Infiltration of CD8⁺ cells in solid tumors including colorectal carcinoma has been correlated with favorable prognosis (Ohtani 2007). In the inflamed colon, Tregs accumulate locally in an effort to limit the inflammation and mitigate the related tissue damage (Holmen et al. 2006). Furthermore, it has been demonstrated that CRC could secrete CCL5 which recruits Tregs to tumors through CCL5/CCR5 signaling (Chang et al. 2012). In fact, CCR5^hiFoxp³⁺ Tregs are more suppressive than CCR5^loFoxP3⁺ Tregs in human CRC (Ward et al. 2015). Tumors can also secrete angiogenic factors such as vascular endothelial growth factor (VEGF) that help recruit Tregs to tumors. Blocking VEGF pathways inhibited the recruitment of FoxP3⁺ Tregs in a mouse model of CRC and impeded the proliferation of human Tregs in vitro (Terme et al. 2013). The high frequency of Tregs might be important for controlling cancer-driving inflammation, but at the same time, they may promote tumor progression by obstructing specific immune responses and contributing to escape of tumor cells from the immune surveillance.

As of now, the state of the field seems to be divided regarding the role of Tregs in the outcome of CRC. Most of the TILs in CRC patients are CD3⁺ and using CD3⁺/FoxP3⁺ T-cell ratio as an appropriate prognostic marker in CRC, a low intraepithelial CD3⁺/FOXP3⁺ T-cell ratio has been associated with shorter patient survival (Sinicrope et al. 2009). Another study used intra-tumoral CD8⁺ T-cell/FoxP3⁺ cell ratio as a predictive marker for overall survival and found FoxP3⁺ Tregs to be associated with lymph node metastases (Suzuki et al. 2010). Similarly, a preliminary study showed that peripheral blood derived CD4⁺CD25⁺ T-cell lines derived from CRC patient and inhibited the proliferation of autologous HLA-A1 restricted CTLs via TGFβ, thus allowing for tumor cells to escape the immunosurveillance (Somasundaram et al. 2002). Moreover, patients with less abundant Foxp3 Tregs displayed improved T-cell responses (Galon et al. 2006).

There have been reports of antigen-specific CD3⁺ T-cell responses in CRC, but spontaneous tumor clearance is a very rare occurrence (Campi et al. 2003; Nagorsen et al. 2005; Sinicrope et al. 2009). Apparently, Treg suppression in CRC is tumor-associated antigen (TAA) specific and not systemic (Betts et al. 2012; Clarke et al. 2006). Another detailed analysis using various CRC-specific TAAs such as carcinoembryonic antigen (CEA) peptides, telomerase, Her-2/neu, and MUC-1 showed that they can lead to Treg activation (Bonertz et al. 2009), which may explain the ineffective immune response against the tumor. Moreover, depleting Tregs in PBMCs from CRC patients markedly enhanced the IFNγ and TNFα production in T cells, which were stimulated with a CEA peptide (Yaqub et al. 2008). TAA-specific Tr1 cells have also been identified using a p53 peptide (Bueter et al. 2006). Highly suppressive TIM3-expressing FoxP3⁺Helios⁺ Tregs have also been demonstrated at different stages of CRC (Toor et al. 2019). Tumor-infiltrating highly suppressive Tregs that express interleukin-1 receptor 2 (IL1R2), PD-1 Ligand 1, PD-1 Ligand 2, TNFRSF9, CCR1and CCR8 have also been reported (De Simone et al. 2016; Fujimoto et al. 2018). Interestingly, high expression in Treg cell signature genes within whole tumor transcriptome, such as LAYN (a transmembrane protein with homology to c-type lectin), and of MAGEH1 (a member of the melanoma antigen gene family) or CCR8, correlated with poor prognosis (De Simone et al. 2016). Another study with CRC patients showed increased levels of Tr1 markers LAG3 and CD49b, as well as IL17 which were inversely associated with patient survival (Chen and Chen 2014). TAA-specific Tregs suppress APC and T-cell function via IL-10 and NK cell function by producing TGFβ, thus incapacitating both adaptive and innate immunity against cancer (Zou 2006). Additionally, tumor-associated Tregs have been shown to express very high levels of CD39 and inhibit the in vitro T-cell transendothelial migration via adenosine synthesis (Sundstrom et al. 2016).

The field is not universally in agreement about the role of Tregs in CRC progression, as the picture becomes more nuanced. Some studies showed favorable outcomes associated with high infiltration with FoxP3⁺ Tregs at different stages of CRC (Frey et al. 2010; Lee et al. 2018; Salama et al. 2009). It is also worth noting that effector Tregs (eTregs) are very frequently found in CRC patients. In an interesting study, Saito and colleagues revealed the heterogeneity of FoxP3 expression in tumor infiltrating Tregs (Saito et al. 2016). The authors showed that CRC tumors with favorable outcomes were actually more infiltrated with Foxp3^loCD45RA⁺ non-Tregs and upregulated inflammatory genes like Il12a, Il12b, Tgfb1, and Tnf. On the other hand, CRC with higher infiltration of FOXP3^hiCD45RA⁻ eTregs resulted in poor prognosis and lower disease-free survival (Saito et al. 2016). Complementing this study, Lin et al. demonstrated that eTregs, but not Foxp3^loCD45RA⁺ non-Tregs, were associated with late-stage CRC (Lin et al. 2013). Another study showed that FoxP3⁺ eTregs that express Blimp1, a critical regulator of CD4 T cell exhaustion, have been identified in human CRC and were associated with long-term disease-free survival (Ward-Hartstonge et al. 2017). FoxP3⁺ Tregs might also be involved in regulating the immune response directed against high-level microsatellite instability (MSI) CRC at the primary tumor site. Significantly elevated FoxP3⁺ Tregs were associated with CRC patients with known MSI status (Michel et al. 2008).

The discrepancies in the role of Tregs may partly be explained by Treg plasticity. Inflammation in the gut and tumor microenvironment drive certain Tregs to display plasticity and an acquisition of a pro-inflammatory phenotype. RORγt⁺FOXP3⁺ Tregs with pathogenic properties have been identified in CRC patients and were enriched in patients with late-stage cancer (Blatner et al. 2012). Interestingly, RORγt⁺FOXP3⁺ Tregs in CRC patients produced both IL10 and IL17 (Blatner et al. 2012), indicating that the balance between protective and pathogenic Tregs could potentially influence the disease outcome. Murine Tregs have also been shown to lose FoxP3 expression and gain the ability to produce IL17, in response to an inflammatory bacterial insult in the gut (Yurchenko et al. 2012). Populations of IL17-producing Tregs have been identified in mouse polyps and human CRC (Blatner et al. 2010). Interestingly, this Treg reprogramming does not represent a classical Th17 conversion as these IL17-producing Tregs did not lose their suppressive function (Blatner et al. 2010). It has been suggested that these Tregs might be transitional and have reserved the ability to differentiate into FoxP3⁺ Tregs or TH17 cells depending on the local milieu of tumor microenvironment (Du et al. 2014). In a mouse model of CRC, TGFβ and prostaglandin E2 mediated the differentiation of Th17 cells into suppressive IL17⁺ and IL17⁻FoxP3⁺ Tregs (Downs-Canner et al. 2017). Indeed, IL17-producing FoxP3⁺ Tregs have been identified in human CRC (Girardin et al. 2013; Kryczek et al. 2011) with the ability to suppress the proliferation of CD8⁺ TILs stimulated with CRC antigens (Ma and Dong 2011).

Besides CD4⁺Foxp3⁺ Tregs, there have been some unconventional Tregs that have been identified in the context of CRC. Populations of FoxP3⁺ and FoxP3⁻ T cells that express latency-associated protein (LAP) have been identified in CRC (Mahalingam et al. 2014; Scurr et al. 2014; Zhong et al. 2017). The tumor-infiltrating LAP⁺FOXP3⁻ cells had much greater suppressive capacity than LAP⁻FoxP3⁺ Tregs isolated from the peripheral blood (Scurr et al. 2014) and were associated with poor prognosis. CD8⁺FoxP3⁺ Tregs have also been identified in CRC tissues (Taylor et al. 2016). Another study demonstrated tumor infiltrating CD8⁺CD25⁺FoxP3⁺ cells to be able to express CTLA4 and GITR and suppress CD4⁺ T-cell proliferation and IFNγ production ex vivo (Chaput et al. 2009).

9.7. Treg-Related Therapeutic Approaches

Based on the observed impairment of Treg immunosuppressive abilities in patients with inflammatory disorders of the gut, therapeutic approaches aiming to correct such deficiencies may have the potential to protect the gut tissue from excessive inflammation and successfully restore intestinal homeostasis. Studies on Treg immunotherapy with infusion of autologous Treg or mesenchymal stem cells (MSCs), or enhancement of Treg numbers and function via administration of antibodies, cytokines, or dietary interventions, have shown preliminary efficacy. In this section, we discuss the available evidence for Treg-targeted therapies to modulate intestinal disorders.

9.7.1. Treg Infusion

Purity, homing ability, antigen-specificity, and survival of Treg cells all need to be carefully considered during the design of therapies involving Treg infusion. Varying in vitro protocols are used to obtain sufficient numbers of Tregs to be used in immunotherapy. The expansion of polyclonal FoxP3⁺ Treg isolated from peripheral blood and the differentiation of Tr1 from human CD4⁺ T in the presence of IL10 and allogenic monocytes are the most common methods employed to obtain Tregs. The use of specific markers to properly identify and isolate Treg is critical to prevent contamination with other cells. Also, to appropriately reach the desired site of action, infused Tregs should express homing receptor(s) that guide their migration to the gut where the cells can act directly at the inflamed site or to the draining lymph nodes to modulate T-cell priming. Engineering or selecting the right antigen specificity of the infused Tregs can influence the expansion of these cells in the gut, where TCR engagement by microbial and/or dietary antigens would promote Treg expansion, retention, and function after transfer.

T cells expanded from the population of blood CD25⁺CD127^lowCD45RA⁺ (natural) Tregs may represent a promising T-cell transfer therapy approach in CD, as they do not convert to Th17, home to the small intestine and suppress the activation of immune cells in the intestinal tissue of CD patients (Canavan et al. 2016). Phase I and II clinical trials to test the safety and efficacy of Treg immunotherapy using cells obtained with this protocol are ongoing (NCT03185000). The upregulation of α4β7 and CCR9, gut homing molecules, and the induction of a transcriptional profile consistent with gut-homing can be enhanced by addition of atRA, usually produced by CD103⁺ cells in the gut, during in vitro expansion of human pediatric thymic and adult blood Tregs (Hoeppli et al. 2019). Additionally, the expansion of Tregs from patients with CD in the presence of RAR568, specific agonist for RA receptor A was even more efficient than atRA in inducing α4β7 (Goldberg et al. 2019). New approaches aim to generate antigen-specific Tregs, which would decrease the risk of a non-specific inhibition, allow the use of lower numbers of Tregs and increase the chance of Treg expansion in the gut. As proof of principle, in a phase I/IIa clinical study in 20 patients with refractory CD, a reduction of 40% in Crohn’s Disease Activity Index (CDAI) was observed after treatment with single-injection of Tr1 ovalbumin-specific Tregs in response to ovalbumin ingestion (Desreumaux et al. 2012).

In patients with GVHD, tTreg infusion was shown to be well tolerated with no toxicities reported and improved clinical outcomes if injected alone or in conjunction with other anti-GVHD therapies (Duggleby et al. 2018). More recently, a study in mice showed that treatment with selective inhibitor of bromodomain and extra-terminal (BET) proteins, which reduce pro-inflammatory cytokine production, can be successfully combined with Treg expansion therapy for the treatment of GVHD (Copsel et al. 2018).

The field of NEC is perhaps less advanced both in the available knowledge of Treg function and their potential use as preventative therapy. One study has reported that administration of Tregs in a mouse model of NEC has attenuated the severity of NEC by limiting the immune response, also indicating the potential of Tregs in controlling human NEC (Dingle et al. 2013). Additionally, in very low birth weight (VLBW) neonates fed with lactoferrin (LF), the major protein in mammalian milk, prevention of NEC has correlated with increased levels of circulating FoxP3⁺ Tregs (Akin et al. 2014).

9.7.2. Promoting the Number and Function of Endogenous Tregs

Indirect interventions to affect the gut Treg population represent viable approaches aimed to maximize Treg immunosuppression in vivo. While majority of the available data comes from preclinical models, some therapies have found their way into ongoing clinical trials and are discussed in this section.

9.7.2.1. Targeting Cytokines and Inflammatory Pathways

Treatments to improve Treg population based on the administration of cytokines known to promote their survival and expansion or inhibition of pro-inflammatory cytokines known to promote Treg dysfunction/cell death have been explored in several studies.

Administration of low doses of IL2 preferentially activates Treg cells due to their abundant expression of CD25, avoiding its unspecific action on other immune cells expressing lower levels of CD25. Low dose of IL2 promotes proliferation, thymic export, and resistance to apoptosis among FoxP3⁺ Tregs (Matsuoka et al. 2013). In a phase I study, patients with active chronic GVHD (cGVHD) refractory to glucocorticoid treatment treated daily with low-dose IL2, augmented Tregs and demonstrated clinical improvement (Koreth et al. 2011). In a phase II trial, 61% of the patients had a clinical response at multiple cGVHD sites, including gastrointestinal tract. During 2 years of extended IL2 therapy, both the clinical response and Treg cell immune response persisted (Koreth et al. 2016). Currently, a clinical trial phase II is recruiting patients to study the induction of Tregs by low-dose IL2 in CD and UC (NCT01988506). Other soluble factor beside IL2 can be used in therapies aimed at Treg expansion. In a mouse model of cGVHD, granulocyte-macrophage colony-stimulating factor (GM-CSF) therapy increased the number CD4⁺CD8⁻ DCs, expanded Tregs, and protected against the development of skin GVHD (Hotta et al. 2019). The potential of GM-CSF therapy for the recovery of intestinal homeostasis is yet to be addressed in future studies.

AMG714, a neutralizing anti-IL15 monoclonal antibody, in a clinical phase IIa study in patients with refractory coeliac disease was safe, and improved the clinical symptoms even though it did not prevent mucosal injury after gluten challenge, particularly diarrhea in patients who underwent gluten challenge (Lahdeaho et al. 2019; Cellier et al. 2019). Although not directly studied in clinical trials, IL15 blockade would be expected to restore TGFβ signaling in the LP immune cells and promote the local Treg population.

Restoring TGFβ sensitivity has also been an active area of research and pharmacologic developments. The oral administration of SMAD7 antisense oligonucleotide, GED-0301 (Morgensen; Celgene), was shown safe and effective in double-blind, placebo-controlled trials (Monteleone et al. 2015, 2016; Feagan et al. 2018), although a more recent clinical trial with CD patients failed to show a benefit over placebo (Sands et al. 2019).

Several studies in mice had shown that therapy that promotes IL33 signaling in FoxP3⁺ Tregs with either rIL33 or sST2-blocking antibody showed the potential to be effective in models of exacerbated intestinal inflammation. The response to such treatments may be contextual, as the intestinal inflammation could be ameliorated if rIL33 was administered during chronic phases of intestinal inflammation (Grobeta et al. 2012), but was exacerbated in acute settings (Oboki et al. 2010). The latter may be attributed to the role of IL33 as alarmin, instigating a pro-inflammatory response in the gut. On the other hand, in chronic stages of the intestinal inflammation, rIL33 could promote tissue homeostasis and healing. In a model of CD, the protection promoted by rIL-33 was mediated by the induction of FoxP3⁺ Treg expansion and CD103⁺ DCs in the gut (Duan et al. 2012). The same way, ST2⁺ Treg expansion during rIL33 pre-conditioning therapy or blocking sST2 during peri-transplant period, decreased GVHD severity and mortality (Matta et al. 2016) (Zhang et al. 2015). One study in patients with celiac disease showed even higher levels of IL33 and sST2 in serum and mucosal samples of celiac patients than in patients with Crohn’s disease, which suggests that the blocking sST2 therapy approach could also have a favorable outcome in celiac disease patients (Lopez-Casado et al. 2017).

Inhibition of inflammatory pathways may also improve Treg survival in the gut. For example, the anti-TNFα therapy has shown to reverse FoxP3⁺ Treg apoptosis in IBD patients (Li et al. 2015). Blocking antibody against OX40L-OX40, promoted Treg responses, such as increased IL10-producing CD4⁺ T cells and the frequency of CD25⁺Foxp3⁺CD127^loCTLA4⁺CCR4⁺ Tregs, reduced T-cell infiltration in various sites of inflammation including the gut, and reduced the severity of the GVHD in immunodeficient mice transplanted with human peripheral blood mononuclear cells (Tripathi et al. 2019).

9.7.2.2. Mesenchymal Stem Cell (MSC) Infusion

Mesenchymal Stem Cells (MSC) are multipotent non-hematopoietic cells present in different tissues, which due to their immunomodulatory characteristics are considered as new therapeutic agents in the cell-based therapy of IBD, NEC, and GVHD. Different tissues were used as the source of MSCs in research studies, including bone marrow, amniotic fluid, umbilical cord, and placenta, where these cells are typically identified based on the expression of markers CD73, CD90, and CD105, and lack of CD34, CD45, CD11b, CD14, CD19, or CD79a (Drucker et al. 2018; Chen et al. 2013). Recently published studies emphasized the crucial importance of MSC for the attenuation of colonic inflammation mainly through the promotion of M2 macrophage lineage (Giri et al. 2020), induction of intestinal epithelial repair (Gong et al. 2016), and the expansion of colonic Treg population (Vicente-Suarez et al. 2015). The effects of MSCs on macrophage-mediated mucosal repair involves polarization of IL10⁺ tissue macrophages via CCL2 and CXCL12 (Giri et al. 2020). In Treg generation, MSCs can either act directly by delivering sphingosine 1-phosphate (S1P) to CD4⁺ T cells and promoting FoxP3-expressing and IL10-producing Tregs (Li et al. 2019). Indirectly, MSCs can promote a regulatory phenotype in DCs, associated with high production of IDO1 (Zhang et al. 2018), suppression of expansion of inflammatory Th1 and Th17 cells, and favoring of conversion to CD4⁺CD25⁺FoxP3⁺ Tregs (Harrell et al. 2018; Zhang et al. 2017).

Cell-free MSC-based therapy have also been postulated as potentially safer. Conditioned medium containing regulatory soluble mediators and exosomes secreted by MSCs mimicked the ability of the MSCs to increase total number of IL10- and TGFβ-producing Tregs in injured tissues and peripheral lymph organs of mice with autoimmune and chronic inflammatory diseases, including IBD (Heidari et al. 2018).

Thus far, MSC infusion in humans has been reported to be safe over long-term use. The efficacy and diversity of the MSC-based therapy protocols being tested in human intestinal disorders have been extensively revised elsewhere (Gregoire et al. 2017; Moheb-Alian et al. 2016; Gao et al. 2016). The improvement in the intestinal disease outcome after MSC therapy is related to improvement in mucosal frequency of FoxP3⁺ Tregs. Allogeneic bone marrow-derived mesenchymal stromal cell (bmMSC) therapy in perianal Crohn’ s disease fistulas resulted in closure at week 24 and remained closed after 4 years (Barnhoorn et al. 2020). Ciccocioppo et al. showed an increased percentage of circulating and mucosal Tregs in CD patients after MSC therapy (Ciccocioppo et al. 2011). The use of MSC as adjuvant therapy combined with conventional pharmacological approaches has also been considered. The combination of MSC with anti-cytokine therapy to treat CD patients with perianal lesions showed significant improvement in comparison to the control group receiving cell therapy only (Knyazev et al. 2018). Neutralizing TNFα and IFNγ during MSC-based therapy may increase their efficacy by removing the ability of these two cytokines to impair the pro-repair effects of MSCs (Hu et al. 2019) and increase Treg generation/viability in the inflamed intestine (Visperas et al. 2014) (Veltkamp et al. 2011).

Several trials in safety and efficacy of MSC therapy are ongoing in both IBD and GHVD (clinicaltrial.gov). The results of this novel therapeutic approach and in-depth information of how its findings relate to Treg improvement in the intestinal lamina propria are awaited by the scientific community.

9.7.2.3. Induction of Tregs by Microbiota

In humans and mice, deviation in gut microbial composition from homeostatic condition, referred to as dysbiosis, is associated with intestinal immune disorders, including NEC (Denning and Prince 2018), celiac disease (Girbovan et al. 2017), and IBD (Ni et al. 2017). In patients with GVHD, the conditioning regimen and medications given to patients undergoing allo-HCT reduce the microbial diversity and disturb the homeostatic crosstalk between the microbiome and the host immune system (Hidaka et al. 2018). However, the direct causal relationship between dysbiosis and human intestinal disorders is not straightforward. While it may be possible that dysbiosis arises as a consequence of the inflammatory milieu in the gut, it is also well established that transfer of dysbiosis community can increase susceptibility to and severity of experimental colitis (Harrison et al. 2018; Gerassy-Vainberg et al. 2018; Garrett et al. 2007). Despite that, research into fecal microbiome transplant (FMT) or next-generation probiotics and preclinical studies with mouse models of intestinal inflammation show promise.

In humans, FMT from healthy individuals to subjects with active CD have successfully increased intestinal microbial diversity and mucosal CD4⁺CD25⁺CD127l^ow Tregs (Vaughn et al. 2016). However, clinical studies of FMT in active IBD, especially CD, brought mixed results, with notable cases of disease exacerbation attributed to the introduction of novel antigens to an already overreactive immune system and/or increasing bacterial load (Sarrabayrouse et al. 2020; Browne and Kelly 2017).

Administration of selected bacterial species or micro-communities derived from the human gut microbiota offers a promising therapeutic alternative to FMT for the prevention and/or treatment of dysregulated intestinal immune responses. Bacterial species from strains clusters IV, XIVa, and XVIII of Clostridia (Narushima et al. 2014), Bacteroides fragilis (Round and Mazmanian 2010) and Akkermansia muciniphila (Zhai et al. 2019) have been reported to promote Tregs in murine colon, as was described in more detail in Sect. 9.4.3. Short-chain fatty acids (SCFAs) such as butyrate and propionate produced by Clostridia through fermentation of mainly undigested dietary carbohydrates are the main mediators of Treg induction by this bacterial population. This aspect of Treg regulation was also discussed in more detail in Sect. 9.4.3. With the notable exception of NEC, where the relative abundance of Clostridia clusters IV, XIVa is increased as compared to healthy infants (Brower-Sinning et al. 2014), these bacterial groups are generally decreased in the intestine under inflammatory conditions such as IBD and celiac disease (Narushima et al. 2014; Sanchez et al. 2010). Their reduction may contribute to intestinal immune dysregulation, and replenishment may offer a therapeutic strategy.

Clostridium F. prausnitzii can induce an immunoregulatory phenotype in human dendritic cells, characterized by the expression of IL10, IL27, CD39, IDO1, and PDL-1 and inhibition of inflammatory mediator production in response to TLR4 stimulation, and ultimately promotes colonic Treg development (Alameddine et al. 2019).

B. fragilis colonization in mice was found to reduce colitis and inflammation-associated colon cancer in mice in a mechanism dependent on the production on PSA and TLR2 signaling (Lee et al. 2018), and promoted circulating Tregs, alleviating intestinal inflammation (Tan et al. 2019). PSA produced by B. fragilis-derived PSA (but no other TRL2 ligands) stimulated FoxP3⁺ Tregs while suppressing intestinal Th17 generation. IL17 suppression by PSA is a critical mechanism by which B. fragilis associates with its host and promotes mucosal colonic niche and immune tolerance (Round et al. 2011). The first study in B. fragilis PSA induction of human Tregs showed that PSA presentation by total APC induced human peripheral CD4⁺ T differentiation into IL10-producing Tr1 cells (Kreisman and Cobb 2011). Later, Telesford et al. showed that PSA promoted frequency and function in human IL10-producing CD39⁺Foxp3⁺ Tregs in the presence of DCs through HLA-DR, CD86, CD40, and PD-L1 and enhanced Treg suppressive function (Telesford et al. 2015). Therefore, the presence and specific phenotype of the APC is required to prime CD4⁺ T cells with PSA to promote differentiation into one of the two Treg subtypes. IBD patients have a significantly lower relative abundance of B. fragilis actively producing PSA, thus not just overall decrease of B. fragilis but also a reduction in PSA production by this species may contribute to the breakdown of the intestinal homeostasis in IBD patients (Blandford et al. 2019).

Akkermansia muciniphila, a mucin-degrading and abundant member of a healthy microbiota, is reduced in IBD (Lopez-Siles et al. 2018; Earley et al. 2019). It is primarily localized in the transverse and descending colon, where its abundance depends on the mucus thickness and is affected by the immune status of the host (Van den Abbeele et al. 2010). A. muciniphila administration in mice modulated the intestinal inflammation and ameliorated acute colitis (Bian et al. 2019). Its beneficial effect in chronic colitis was associated with an induction of Treg differentiation (Zhai et al. 2019). The role of A. muciniphila in preventing intestinal disorders has to be addressed with caution because in contrast to the findings in colitis, expansion of A. muciniphila in mice with GVHD was associated with loss of the colonic mucus layer and compromised intestinal barrier function (Shono et al. 2016).

There is an accumulating evidence that restoration of intestinal flora by classical probiotic bacteria also generates FoxP3 Tregs in the gut, making the probiotic consumption an interesting approach to induce this cell population in patients with intestinal disorders. Commonly probiotic mixture contains Bifidobacterium spp., Lactobacillus spp., and Streptococcus thermophilus. The beneficial impact of probiotics on intestinal homeostasis has been shown in many studies (Abraham and Quigley 2017; Cristofori et al. 2018) (Patel and Underwood 2018; Andermann et al. 2016). For instance, L. fermentum, L. acidophilus, and L. casei Lbs2 protect from experimental colitis, through the induction of Treg responses (Jang et al. 2019; Park et al. 2018; Thakur et al. 2016). Treg promotion by Lactobacilli is usually mediated by induction of a regulatory phenotype in the intestinal DCs (Mikulic et al. 2017). Furthermore, Bifidobacteria and S. thermophilus stimulated significant concentrations of TGFβ in human PBMC (Donkor et al. 2012). Oral consumption of Bifidobacterium infantis induces FoxP3 expression and enhances IL10 and TGFβ production in the mouse colon (Zhou et al. 2019). Although limited clinical trials suggest a promising efficacy of classical probiotic strains and their combinations in ulcerative colitis, there remains a lot of skepticism in the IBD field regarding the efficacy of such approaches in Crohn’s disease (Guandalini and Sansotta 2019).

9.7.2.4. Dietary Vitamins

Vitamins are organic dietary compounds with significant impact on the gut mucosal immune function. As we described in more detail in Sect. 9.4.4., vitamins A and D provide important signaling cues for the generation and survival of intestinal Foxp3⁺ Tregs (Huang et al. 2018; Fakhoury et al. 2020). Retinoic acid (RA), product of vit A degradation by RALDH2 expressed by CD103⁺ DCs in the intestine, has a large spectrum of modulatory effects on Tregs. Such pivotal role of vitamin A metabolites on intestinal Foxp3⁺ Tregs and homeostasis is illustrated by the findings that vitamin A deficiency exacerbates experimental colitis (Reifen et al. 2002) and supplementation of vitamin A results in decreased inflammatory markers in NEC patients (Xiao et al. 2018). Therefore, administration of vitamin A/retinoic acid in clinical patients with outcomes related to intestinal inflammation may stimulate FoxP3⁺ Tregs. A clinical trial on the efficacy of high doses of vitamin A to prevent gastrointestinal GVHD out of the Ohio State University Comprehensive Cancer Center is in early stages (NCT03719092), and the outcome of this therapeutic approach may inform other intestinal disorders going forward.

Along with vitamin A, vitamin D is another potent immunomodulator that affects T cells directly and indirectly (Cantorna et al. 2015, 2019). We have already discussed the effects of vitamin D on Treg biology earlier in this chapter. In IBD patients, VDR polymorphisms have been associated with higher disease risk and increased morbidity (Eloranta et al. 2011). Vitamin D deficiency is commonly observed in IBD patients, and association between vitamin D status and the extent of inflammation has been reported by several studies (Jorgensen et al. 2013; Gubatan et al. 2019; Fletcher et al. 2019; Burrelli Scotti et al. 2019; Law et al. 2019). In a double-blind randomized controlled trial, where 90 mild-to-moderate UC patients were administered one single muscular injection of vitamin D (7.5 mg) and reduced serum concentrations of TNFα, IFNγ, and IL12p70 levels, but it did not affect serum IL4 and IL10 levels. However, no other clinical parameters were evaluated or included in the report (Sharifi et al. 2019). In a murine model of GVHD, vitamin D promoted FoxP3⁺ Treg generation and slowed the disease progression (Ni et al. 2019). A severe vitamin D deficiency has also been documented in patients with celiac disease (Sulimani 2019) and maternal vitamin D deficiency has been proposed to be a new risk factor for NEC in preterm infants (Cetinkaya et al. 2017). Importantly, vitamin D has pleiotropic effects on the cells of the innate and adaptive immune system and affects intestinal barrier function and gut microbial ecology, thus any beneficial effects of this vitamin/hormone cannot be solely attributed to promotion of Tregs and their immunosuppressive functions.

Finally, dietary plant sterols capable of reducing intestinal inflammation in mice also promote Treg numbers. These compounds are consumed in habitual diets derived from vegetable oils, grain products, nuts, seeds, fruits, and vegetables (te Velde et al. 2015). Fiber-enriched diets were associated with a higher colonic in situ production of total SCFA including butyrate, which promotes the gut Tregs (Rodriguez-Cabezas et al. 2002). Dietary interventions and application of prebiotics that enhance butyrate production, such as 2′-fucosyllactose, may help induce remission in UC without involving additional immune suppression (phase IV clinical trial NCT02345733) or may contribute to the maintenance of remission in pediatric and young adult IBD patients undergoing anti-cytokine therapy (NCT03847467). Collectively, the available data suggests somewhat instinctively that the supplementation with vitamins (or correction of deficiencies) and intake of a balanced diet rich in fruits, grains, and vegetables should be a part of the comprehensive therapy of intestinal inflammatory disorders, and a part of the associated mechanism may be through promotion of intestinal Tregs and immunoregulatory environment in the gut.

9.8. Conclusions

Intestinal Tregs is a complex and heterogenous population of immunosuppressive cells whose development and function are affected (positively and negatively) by an array of factors, both host-intrinsic and extrinsic. They play crucial roles in maintaining the low mucosal inflammatory tone in the gut under continuous antigenic challenge, contribute to the maintenance of intestinal barrier function, keep the immune response to pathogens in check, and promote the resolution of inflammation and mucosal restitution. As such, they represent attractive target of research and therapeutic developments. Although much is known already about their phenotypic and functional features in mice, translating these results to humans has been challenging. A better understanding of human Tregs at mucosal sites, especially in the gut, will help guide better therapeutic options, primary or adjuvant, to treat chronic inflammatory conditions of the gut.