Mechanisms of tolerance and allergic sensitization in the airways and the lungs

Abstract

The respiratory mucosa is constantly exposed to non-infectious substances that have the potential of triggering inflammation. While many particles are excluded, soluble molecules can reach the epithelium surface, where they can be uptaken by dendritic cells and stimulate an adaptive immune response. Most mucosal responses result in tolerance to subsequent antigen encounters, which is mediated by Foxp3⁺ regulatory Tcells. Genetic and environmental factors, added to the ability of certain allergens to induce innate responses, can predispose to allergic sensitization. In this review we discuss recent advances in the understanding of the mechanisms of tolerance and allergic sensitization to airborne allergens.

Introduction

In normal conditions, inhalation of non-infectious protein antigens induces a state of tolerance, which is essential to maintain mucosal homeostasis under constant exposure to environmental antigens. Studies of respiratory tolerance in rodents have provided valuable information on the mechanisms that prevent allergic sensitization through the airways. Resident lung cells set the stage for a non-inflammatory steady state. Airway macrophage interactions with epithelial cells involving integrins and TGF-β are necessary to prevent inflammation in the lung. Antigen-presentation by migratory lung DC to CD4⁺ cells in draining LN and induction of Foxp3⁺ iTreg cells are essential steps in the establishment of tolerance. Allergic sensitization occurs by a failure in the mechanisms of tolerance, leading to development of inflammatory airway diseases such as rhinitis and asthma. Allergies are caused by Th2 lymphocytes that secrete IL-4, IL-5 and IL-13, and by IgE antibodies. In this review we will focus on recent data on airways immune tolerance and sensitization.

Induction of tolerance

Antigens entering through the respiratory mucosa normally induce tolerance

Exposure to environmental allergens does not elicit clinical symptoms in non-allergic individuals. Human studies of cord blood and neonatal responses to allergens indicate that this is not due to absence of immune response, but rather to development of a protective response. Experiments of inhalation of allergens in rodents by the Patrick Holt group in the 1980s and 1990s clearly established that inhaled antigens induce a state of dominant tolerance that could be transferred by T cells. These days, the nomenclature dominant tolerance implies the activity of regulatory T cells. Similarly to oral tolerance, tolerance to antigens entering through the respiratory tract is antigen-specific and systemic [1].

Adaptive (or inducible) Regulatory T cells

It has been shown that several strategies leading to tolerance induction do so by inducing regulatory cells. Although not all regulatory T cells express Foxp3 [2,3], absence of Foxp3 compromises immune tolerance leading to severe autoimmunity both in animal models [4] and human patients [5]. Foxp3⁺ regulatory T cells are generated either in the thymus (nTreg cells) or induced in the periphery (iTreg cells) [6]

The study of peripheral Treg cell conversion was greatly facilitated with TCR-transgenic RAG deficient mice, which are unable to produce a functional TCR from their endogenous genes and, as a consequence, lack nTreg cells [6]. Therefore, it becomes possible to examine the conversion of such T cells into Foxp3⁺ Treg cells, in the absence of a possible contamination due to the expansion of pre-existing nTreg cells.

It was shown that addition of TGF-β to in vitro cultures of OVA-specific T cells from DO11.10 mice led to induction of Foxp3⁺ Treg cells [7]. Furthermore, the use of low concentration of OVA peptide for in vitro activation of Foxp3⁻ T cells also resulted in induction of Foxp3⁺ T cells [8]. Interestingly, B cells were shown to be more efficient than DCs in driving Treg conversion, presumably by their inability to provide full co-stimulatory signals [8].

There are now many examples of in vivo generation of iTreg cells from naïve T cells from TCR transgenic RAG-deficient mice. Oral and nasal exposure to OVA was shown to lead to the conversion of TCR-transgenic RAG-deficient T cells into Foxp3⁺ iTreg cells, a process that was dependent on TGF-β [9]. Furthermore, mucosal tolerance (both oral and nasal) was completely abolished if Foxp3⁺ iTreg generation was prevented [10].

Exposure of T cells to a low dose of persistent antigen also resulted in Foxp3⁺ iTreg induction [11], as did the administration of antigen covalently coupled with anti DEC205 antibody [12]. It was also possible to document the in vivo induction of iTreg cells in TCR-transgenic RAG^−/− mice specific to a given male following transplantation of male skin grafts onto female TCR-transgenic mice treated with tolerogenic anti-CD4 MAbs [13]. Immature DCs or DCs treated with vitD3 (a reagent that prevents subsequent DC maturation) when adoptively transferred into male-specific TCR transgenic female mice equally lead to peripheral induction of Foxp3⁺ Treg cells and dominant tolerance to male skin grafts [14]. The peripheral induction of Foxp3⁺ iTreg cells, both in vitro and in vivo, requires TGF-β as it is abrogated in the presence of neutralizing anti-TGF-β MAbs [9,13,14].

Using TCR transgenic T cells specific to a moth cytocrome c (MCC) peptide, it was possible to show that induction of peripheral Foxp3 expression was a consequence of activation by a low dose of a strong agonist, while weak agonists only induce a transient population of Foxp3 expressing cells [15].

Spontaneous conversion of non-regulatory T cells into Foxp3⁺ Tregs was also shown following adoptive transfer experiments of CD4⁺CD25⁻ T cells into congenic mice [16,17]. However, in animals that are not TCR-transgenic RAG^−/− it is always difficult to exclude a contribution of Treg expansion by Foxp3⁺CD25⁻ cells.

The anatomical location where Treg cells operate in vivo is important. Treg cells able to prevent transplant rejection can be found infiltrating the tolerated transplant [18]. Remarkably, Treg cells accumulate preferentially within the tolerated graft, e.g. transplanted skin, but not within the host-origin skin. It will be important therefore to clarify whether a local immune privileged-like site may be created through the action of Treg cells. Such Treg-induced immune privileged sites may be important for the generation of more iTreg cells (“infectious” tolerance)[19].

In mice displaying normal T cell repertoire, tolerance to inhaled OVA was demonstrated to depend on regulatory T cells expressing TGF-β on their surface [20]. For the reasons discussed above, it was not possible to determine whether these TGF-β⁺ cells were derived from nTreg cells or de novo-induced iTreg cells. It was however possible to establish that TGF-β⁺ Treg cells expressed Delta-like and Jagged Notch ligands and were thus able to activate the Notch1 expressed on T cells. Reversal of Notch1 activation with γ-secretase inhibitor blocked the suppression brought about by TGF-β⁺ Treg cells [21]. Thus, notch signaling is involved in Treg-mediated suppression of allergic inflammation. However, as will be discussed below, Notch signaling delivered by Jagged-2⁺ DC is involved in Th2 cell induction. Naturally, the presence of modifying factors on the Tregs versus DC must be critical in determining the suppressed versus the Th2 fate, respectively.

A different population of regulatory T cells, named Tr1 cells, have been described [22]. These cells do not express Foxp3, are peripherally induced by antigenic stimulation in an IL-10 rich environment, and are characterized by IL-10 production. It was recently shown that triggering of the aryl hydrocarbon receptor (AhR) can promote the differentiation of human Tr1 cells, while AhR activation in presence of TGF-β induces Foxp3⁺ Treg cells [23]. Furthermore, a synergy between c-Maf and AhR signaling was identified for the induction of Tr1 cells [24].

IL-10-producing Treg cells with potent suppressive ability upon transfer were generated in the lung-draining bronchial lymph nodes by exposure to antigen delivered intranasally, in a process that required ICOS-ICOSL interactions [25].

The hygiene hypothesis and iTreg cells

Epidemiological observations revealed that the prevalence of allergic sensitization was lower in children growing up in a farm environment than in children from urban dwellings. Other factors that were found to be associated with protection from allergic sensitization in childhood were having older siblings and being exposed to domestic animals [26].

The hygiene hypothesis explained these observations by linking the exposure to a richer microbial environment to protection from allergic sensitization. Initially, it was thought that these protective factors lead to development of Th1 response, thus that protection involved a mechanism of immune deviation. The fact that autoimmune diseases, many of which are mediated by Th1 or Th17-biased responses, have also become more prevalent in recent decades, argues against immune-deviation as the main mechanism driven by microbial environment products.

A more general defect in the maturation of the regulatory T cell compartment is likely. Foxp3⁺ Treg cells are dependent on activated effector cells as their main source of IL-2, which is required for Treg cell homeostasis and function. A low T effector activity during early childhood may result in an underdeveloped Treg cell compartment. Moreover, microbial antigens have an effect on the number and maturation of mucosal DC. These, in turn, are essential for Foxp3⁺ iTreg cell induction. A recent report has shown that gut microbiota can induce local Treg development [27]. The human commensal B. fragilis colonization of the mouse gut, or one of its immunomodulatory molecules (polysaccharide A) can induce Foxp3 expression by T cells, endowing them with IL-10-producing ability. Of note, oral exposure to a potent NKT cell agonist – α-galacosylceramide – was shown to lead to induction of Foxp3 expression by NKT cells in the gut. This process was TGF-β-dependent as no Foxp3 induction occurs in mice with NKT cells devoid of a functional TGF-β receptor [28]. All these data are consistent with the observation that germ-free mice are impaired in their ability to establish long-term oral tolerance unless exposed to bacterial products [29]. As mentioned, mucosal tolerance is dependent on Foxp3⁺ iTreg cells [10].

Dendritic cell, macrophages, and other cells inducing tolerance or preventing tissue damage caused by respiratory antigens

Tolerogenic DC [25], alveolar macrophages [30] and other presenting cell types have been implicated in the induction of tolerance in the lungs and airways.

The place of induction of the Foxp3⁺ cells required for immune tolerance in the respiratory tract [10] is not known. However, it is known that it is dependent on lung dendritic cells migrating to the draining bronchial lymph nodes, a process that is dependent on CCR7 [31]. CD103⁺ gut DC were shown to be important in the generation of iTreg cells [32,33]; these cells migrate to the gut-draining mesenteric lymph nodes. Although both CD103⁺ and CD103⁻ lung DC migrated to the draining lymph nodes [31], the migration requirement to establish mucosal tolerance in the lung resembles the requirement for oral tolerance, which requires intact mesenteric lymph nodes.

In the lung-draining lymph nodes, a comparison of dendritic cell populations between mice subjected to tolerance versus inflammation conditions show the predominance of cells with markers of plasmacytoid DC in the tolerant mice; in contrast, CD11c⁺CD11b⁺, MHC class II^high conventional DC predominated in the inflammatory condition [34].

Recently, a population of myeloid-derived suppressor cells induced by high dose LPS was identified. These cells could suppress lung dendritic cells’ ability to reactivate Th2 cells locally. This observation may link the known beneficial effect of LPS on the incidence of allergic disease with the cell type capable of suppressing the Th2 response [35].

In the lungs, a delicate balance must exist in relation to the generation of active TGF-β. As noted above, all studies indicate that TGF-β is required for Foxp3⁺ iTreg cell development, and that iTreg cell development is essential to establish mucosal tolerance. On the flip side, excess TGF-β is associated with lung fibrosis. In the lungs, TGF-β is locally activated by cells that express the integrin alpha v beta 6. As expected, mice lacking alpha v beta 6 develop exaggerated inflammation, because TGF-β activation is impaired; however, these mice are protected from pulmonary fibrosis [36].

Tissues have other means to prevent damage. A key enzyme involved in tissue protection is HO-1 [37]. HO-1 catalyses heme degradation leading to the local production of equimolar amounts of carbon monoxide (CO), biliverdin, and free-iron, which induces the expression of heavy chain ferritin, an iron-binding protein. All three metabolites resulting from heme degradation by HO-1 may have an immune protective effect [38].

Furthermore, the local immune response seems to be under tight control of local availability of specific essential amino acids [39]. The tryptophan-catabolizing enzyme indoleamine 2,3-deoxygenase (IDO) was proposed to be a key component of fetal tolerance [40]. IDO was also implicated in the maintenance of tolerance following co-stimulation blockade by CTLA4Ig, leading to Treg cell induction [41]. These observations were recently generalized with the finding that several other enzymes that catabolize essential amino acids are over-expressed under tolerogenic conditions [39].

Allergic sensitization

Allergic sensitization: IL-4 and IL-13

IL-4 and IL-13 are the most important cytokines for the induction of allergic asthma. IL-4 can directly induce differentiation of Th2 cells that produce an array of inflammatory cytokines, including IL-4 itself, IL-5 and IL-13. Although differentiation of Th2 cells can be induced in the absence of IL-4, IL-4 greatly amplifies the response. IL-4 is the main cytokine involved in IgE production, and through Th2 amplification, in eosinophilic inflammation. IL-13 is mostly responsible for the effects on lung tissue, such as the bronchial hyperresponsiveness, overproduction of mucus by epithelial cells, thickening of smooth muscle and subepithelial fibrosis (reviewed in [42].

Polymorphisms in IL-4, IL-13, or their receptors, that result in higher production of the cytokines, or increased signaling through STAT6, are associated with increase risk of asthma [43]. Whether a genetic predisposition toward higher production of Th2 cytokines or increased Th2 cytokine signaling can by itself override induction of tolerance to innocuous antigens is not clear. Most likely other genetic and environmental factors, as well as the ability of particular allergens to stimulate innate immunity, are required to break tolerance [44]. For example, mice of the BALB/c strain are genetically predisposed to mount Th2 responses [45], yet they can be become tolerant if they are administered chicken ovalbumin (OVA) though the oral or respiratory routes. Th2 cytokines are not induced during this tolerization process [10], and this explains why these mice become tolerant even though they are prone to Th2 responses. However, co-administration of low doses of LPS [46], or protease allergens [47] through the airways, would induce allergic sensitization to OVA.

Early respiratory infections in children have been associated with later development of asthma [48]. Enhanced allergic sensitization after respiratory virus infections can also be observed in experimental systems. Kim and collaborators found that after the clearance of respiratory virus infection in mice, a population of NKT-stimulated IL-13-producing macrophages persisted in the lungs and facilitated allergic sensitization [49].

The interplay between IL-4 and TGF-β is highly relevant. As indicated above, naïve cells can be induced to differentiate into Foxp3⁺ iTreg cells by TCR stimulation in the presence of TGF-β. Interestingly, Foxp3 expression on differentiated Th2 cells silences the IL-4 locus [50]. On the other hand, IL-4 inhibits TGF-β-induced iTreg differentiation in vitro, promoting instead the produciton of IL-9 and IL-10 by T cells [50,51]. Thus, if IL-4 is present in vivo during the initial response to a mucosal antigen, the induction of iTregs could be suppressed and Th2 differentiation can occur.

Upstream of IL-4 and IL-13: Notch, TSLP, IL25 and IL-33

Stimulation of naïve T cells in the presence of IL-4 induces STAT6-dependent IL-4R signaling and differentiation of Th2 cells. However, signals other than IL-4 can initiate Th2 differentiation [52]. For example, DC incubated with parasite extracts, cholera toxin or PGE2 can induce Th2 responses independently of IL-4 [53,54]. Under these Th2-inducing conditions the Notch ligands Jagged1 or Jagged2 are expressed in APC, and there is strong evidence demonstrating that Jagged expression in DC promotes Th2 differentiation, and that Notch expression by T cells is required for Th2 differentiation [55,56].

In recent year it has become apparent the cytokines TSLP, IL-33 and IL-25 (IL-17E), all of which can be produced by epithelial cells, can initiate Th2 responses in the airways and probably contribute to the amplification of allergic responses [44,57-61]. TSLP is an epithelial cell-derived cytokine that can strongly activate myeloid DC. TSLP activated-DC promote differentiation of Th2 cells, expansion of memory Th2 cells and enhancement of the production of IL-5 and IL-13 through OX40/OX40 ligand interactions [60]. TSLP-DC also augments expression of the IL-25R (IL-17RB) in Th2 memory cells but not other cells.

IL-33 is a member of the IL-1 family that was identified as the ligand for ST2 (IL-1RL1), a receptor stably expressed in Th2 cells. IL-33 is expressed by non-hematopoietic tissue cells such as fibroblasts, epithelial and endothelial cells, localizes to the cell nucleus, can be released by necrosis but is inactivated by apoptosis [62,63]. IL-33 is thus consider an “alarmin”. It was recently demonstrated that, in contrast to IL-1 and IL18, IL-33 is biologically active independently of caspase-1 cleavage [64]. The regulation of IL-33 release and the multiple effects of IL-33 are not yet fully understood. IL-33 is a potent inducer of Th2 responses but has also been associated with non-atopic diseases [61].

IL-25 was initially identified as a cytokine produced by differentiated Th2 cells but not naïve cells. In untreated wild type mice, low levels of IL-25 transcripts were detected in the gastrointestinal tract and the uterus but not in lymphoid organs. Subsequently, it was found that in addition to Th2 cells, IL-25 could also be produced by IgE-stimulated mast cells, alveolar macrophages and epithelial cells stimulated by allergens [57].

Infusion of IL-25 into mice stimulated the production of IL-4, IL-5, IL-13 and IgE, and induced allergic-type pathology in the lung. Tissue pathology was dependent on IL-13 but not IL-4 production, and could be induced in the absence of T and B lymphocytes [65]. In this study, the IL-25-responsive, IL-13 producing population was identified as MHC II^hi CD11c^lo and lineage negative. Interestingly, IL-25-induced hyperresponsiveness to the broncho-constrictor methacholine in mice deficient for typical Th2 cytokines IL-4, IL-5, IL-9 and IL-13 [66].

Following the initial observations of Fort and collaborators on the IL-13 producing IL-25-responding cells, several groups have further characterized nonB-nonT cell populations considered to be of the “innate immune system” that respond to IL-25 and IL-33 with production of IL-13 and other Th2 cytokines, and that may be involved in initiating allergic responses. Lineage negative cells in gut-associated lymphoid tissue were found to be Sca-1⁺c-Kit⁺, to produce Th2 cytokines in response to IL25 and IL-33 and to expand after helminth infection or cytokine treatment [67–70]. While these cells were shown to be important for the immune response resulting in clearance of helminth parasites, their role in sensitization to aero-allergens allergens in not known. After allergic sensitization, or in response to intranasal administration of IL-25, a population of IL-17RB⁺iNKT was detected that could induce airway hyper responsiveness upon adoptive transfer [71].

Conclusions

The last years have seen a tremendous increase in the understanding of the immune responses to allergens. Advances in the regulatory T cell field lead to the discovery of peripheral induced Tregs and their importance in tolerance and homeostasis at mucosal tissues. A considerable amount of work has unveiled innate immune mechanisms and cells linked to the induction of Th2 cytokines. There has also been progress in understanding how changes in the environment and life style have lead in recent decades to an increase in allergic diseases, as well as the nature of gene-environment interactions in promoting allergy. The next challenge is to integrate the knowledge of these diverse fields to be able to make predictions as to which individuals have the highest likelihood of develop allergies, and to develop strategies to blunt the increased incidence of allergic inflammation. From the population point of view the challenge ahead is to device strategies that promote tolerance and avoid sensitization at early age.