Food allergy: mechanisms and therapeutics

The immunologic mechanisms responsible for the development of allergic sensitization rather than tolerance to foods are not well understood, although there have been a number of recent advances in our understanding of why some foods are inherently allergenic. In addition, the involvement of alternative routes of exposure that are not inherently tolerogenic may play a role in sensitization to foods. Although there are no currently accepted therapeutic approaches to food allergy, there are a number of approaches to treatment in preclinical or clinical trials. Here, we review selected findings published since 2009 that advance our understanding of mechanisms and new therapeutics for IgE-mediated food allergy.

Introduction

Food allergy is defined as an immunologically mediated adverse reaction to foods, and as such encompasses a range of disorders including IgE-mediated anaphylaxis, food protein-induced enterocolitis syndrome, and food-induced eosinophilic gastrointestinal disorders. For the purpose of this summary of recent advances in the field, we will focus on mechanisms and emerging treatments for IgE-mediated food allergies. Readers are referred to the recently published NIAID guidelines on food allergy for a discussion of topics not covered in this update [1].

Prior exposure to a food antigen by the oral route generates a regulatory T cell response that can then suppress allergic sensitization to that food allergen. There is a lack of consensus about the phenotype of regulatory T cells that prevent food allergy. Hadis et al. [2^•] recently showed that oral tolerance could suppress experimental food allergy through the development of antigen-specific Foxp3⁺ T cells. This was shown definitively using ‘DEREG’ mice that express the diphtheria toxin (DT) receptor under the Foxp3 promoter. Specific ablation of Foxp3⁺ T cells with DT after antigen feeding abolished oral tolerance. In humans, antigen-specific CD25⁺ Foxp3⁺ Tregs are associated with the onset of clinical tolerance to milk [3].

Tolerance is initiated by dendritic cells (DCs) residing in the gastrointestinal lamina propria. CD103⁺ DCs capture antigen in the lamina propria, migrate, and initiate oral tolerance in the draining lymph node by activation of antigen-specific Tregs that then migrate back to the lamina propria. CX₃CR1⁺ DCs/macrophages that are resident in the lamina propria expand the pool of antigen-specific Tregs that can then suppress food allergy [2^•]. Modification of food antigens by adding sugar structures that allow binding to the receptor SIGNR1 on gastrointestinal DCs enhances tolerance through induction of IL-10-producing Tregs [4^•], presenting a potential future approach for immunotherapy. It is not yet known if Tregs can be used therapeutically once sensitization has already been established.

In order to generate allergic sensitization to foods experimentally, adjuvants are commonly used to break oral tolerance. Emerging data suggest that allergic sensitization may occur if the naturally tolerogenic oral route is not the primary route of exposure. Household exposure to peanut has been shown to be associated with allergic sensitization to peanut in children, independent of maternal ingestion [5]. One important route of sensitization may be the skin. Supporting this hypothesis, loss-of-function mutations within the filaggrin gene were found to be associated with peanut allergy independent of atopic dermatitis [6^•]. The filaggrin gene encodes the skin epidermal protein profilaggrin that contributes to barrier function of the skin. Mice deficient in filaggrin are susceptible to allergic sensitization through the skin [7]. The allergenic potential of the skin as a route of exposure is highlighted by the ability to sensitize mice to food allergens via the skin in the absence of adjuvant [8]. However, against the conclusion that the skin is inherently allergenic is the finding that tolerance can also be induced via skin exposure [9]. Furthermore, other relevant allergens such as milk α-lactalbumin require exogenous adjuvant to generate productive sensitization through the skin by promoting antigen presentation by dermal DCs [10]. The different capacity of food allergens to induce adjuvant-independent sensitization via the skin may be due to differences in activity on the innate immune system, as discussed below.

The majority of food allergic reactions are induced by a limited number of food allergens. Accumulating data suggest that activation of the innate immune system is a property of common food allergens. Confirming earlier findings with human DCs, peanut was shown to alter the phenotype of mouse DCs independent of TLR signaling [11]. Peanut and cashew extract, but not milk or egg allergens, can induce anaphylaxis in naïve mice primed for anaphylaxis through pretreatment with propranolol and IL-4 [12^•]. This was triggered by activation of the complement pathways and downstream activation of macrophages. These data show direct innate effects of nut extracts. Milk contains sphingolipids that have recently been shown to activate invariant NKT cells and induce production of Th2 cytokines from the responder cells [13]. Food processing can also enhance innate activity of food allergens. Generation of advanced glycation endproducts during processing of a model food allergen (ovalbumin) resulted in enhanced uptake and presentation by human and mouse DCs by enhanced binding to scavenger receptors [14, 15]. In addition to innate factors that promote sensitization, foods can also contain factors that suppress sensitization. Isoflavones found in soy are directly suppressive on gastrointestinal DCs [16], which may explain why soy is a weak food allergen despite homology to peanut allergens. Innate immune modulatory actions of food allergens are summarized in Figure 1.

Innate activity of allergens does not explain why only some individuals become sensitized to foods. Gastrointestinal epithelial cells at the interface between the gastrointestinal contents and the mucosal immune system are host factors that likely determine the immune response to foods. Epithelial cells from food allergic subjects express higher levels of galectin-9 that can act on DCs to promote allergic sensitization [17]. The epithelial cytokine thymic stromal lymphopoietin (TSLP) is critical for gastrointestinal but not systemic manifestations of food allergy in mice [18]. Mutations resulting in enhanced expression of TSLP are associated with eosinophilic esophagitis [19], but the relationship to IgE-mediated food allergy has not yet been addressed in humans.

Central to the pathophysiology of food allergy is the generation of food-specific IgE. Class-switching to IgE is supported by T cell production of IL-4 and IL-13. Short-term (six hours) stimulation of human peripheral blood mononuclear cells (PBMCs) with peanut or tetramer staining of freshly isolated human PBMCs has been used to phenotype the allergen-specific T cell response [20, 21•]. Both studies emphasized that the frequency of allergen-specific T cells was a magnitude higher in peanut-allergic individuals than healthy controls. Interestingly, peanut-specific T cells expressed CCR4 (consistent with skin homing) but not β7 (associated with gut homing) [21^•]. Changes in production or responsiveness to Th2 cytokines may underlie individual susceptibility to food allergy. A gain-of-function mutation in the IL-4 receptor was shown to result in increased susceptibility to allergic sensitization to foods in mice [22]. In addition to IL-4, IL-9 and IL-13 are critical for gastrointestinal manifestations of food allergy [23, 24], potentially through direct action on gastrointestinal epithelial cells [25].

IgE-mediated food allergy is believed to result from triggering of mast cells to release histamine that acts on target cells including endothelial cells, epithelial cells, and smooth muscle. Studies in mouse models have identified mast cell-derived platelet activating factor as another important mediator of anaphylaxis [26]. Alternative pathways of anaphylaxis, involving IgG and macrophages, can also participate in peanut-induced anaphylaxis in mice [27]. Immunoglobulin free light-chains have also been shown to participate in casein-triggered hypersensitivity reactions in the skin, by an as-yet-unidentified effector mechanism [28]. The contributions of these alternative mechanisms to food-induced anaphylaxis in humans have not yet been determined. Human studies have shown that mutations in the NLRP3 gene that result in either enhanced transcription or stability are associated with food and aspirin-induced anaphylaxis, but not food sensitization [29]. Mechanistic studies explaining the contribution of NLRP3 or inflammasome signaling to anaphylaxis have not yet been performed, but may reveal the existence of novel mechanisms of anaphylaxis to food. Figure 2 summarizes the findings to date on the mechanisms of food-induced anaphylaxis.