Key points

Introduction

Food allergy is a growing public health problem that is estimated to affect 4% to 8% of children and 5% of adults. Although it is not currently understood why some individuals develop allergic sensitization whereas others have immunologic tolerance, an increasing body of evidence is helping to elucidate many of the elements involved in the development of food allergy. In this review, we discuss our current understanding of the pathophysiology of food allergy, from oral tolerance, to sensitization, and last the elicitation of an allergic response. As much of the existing evidence for the mechanisms of food allergy is derived from animal models, we include these studies where relevant. In addition, whenever possible, we review similar evidence involved in human disease and provide applications for consideration in clinical practice.

Oral tolerance

Our immune system evolved to protect us against infection, and one task essential to this defense is differentiating friend from foe. Although much of immunology focuses on self and nonself, within the immunology of allergy it seems clear that healthy individuals are constantly being exposed to nonself proteins yet do not elicit pathogenic responses, whereas those with allergies do. The enormity of this task is arguably most crucial at the epithelial and mucosal surfaces such as the skin, respiratory tract, and gastrointestinal tract, where the body encounters a number of proteins, both friendly and potentially pathogenic.

The gastrointestinal (GI) tract comprises an estimated 300 square meters of surface area. This vast territory presents a tremendous challenge to the innate surveillance systems of the immune system. The challenge arises not only in the sheer quantity of space that the GI tract provides for potential pathogens and antigens to breach its barriers, but also in the volume of proteins that pass through it on a daily basis. The GI tract encounters an estimated 70 to 100 g of protein every day. The mucosal immune system has evolved sophisticated mechanisms by which it can determine when an immune response is warranted, as well as ways to maintain control over aberrant responses, and the balance between these mechanisms is critical to health.

Oral tolerance is defined as the systemic suppression of cellular and humoral immune responses to an antigen that has already been encountered via the oral route. There are a number of factors thought to contribute to the GI tract’s ability to develop oral tolerance, and these include physical barriers, the digestive process, specific immune cells, and immune regulation.

The barrier function of the GI tract is aided by the presence of a protective, hydrophobic mucus-coated surface and of the secretory immunoglobulin, IgA. The mucin oligosaccharide layer helps trap antigen, whereas secretory IgA binds food proteins and prevents absorption of antigen across the intestinal epithelium. One study investigating whether mucosal IgA antibodies protect against food allergy demonstrated that mice that were deficient in the polymeric immunoglobulin receptor, which secretes IgA into the intestinal lumen, were hypersensitive to IgG-mediated anaphylaxis, suggesting that mucosal IgA plays an important part in protecting against food antigen-directed immunogenicity. Strait and colleagues, however, suggested that systemic IgA may be more important than secretory IgA in the GI tract for protection against IgE-mediated anaphylaxis.

The digestive process also plays a role in the generation of oral tolerance. Gastric acid and digestive enzymes help break down food proteins for nutrient absorption. This process also reduces some of the linear and conformational epitopes of food proteins into less immunogenic chains of dipeptides and tripeptides. Acid is also likely to play a role, because clinically, use of acid-neutralizing therapies for ulcer treatment has been demonstrated to drive IgE production against dietary allergens. Further work in animal studies has shown changes in the gastric pH are important for this IgE production and lead to changes within the intestinal epithelium of the mice that are similar to those seen in human food allergy.

Three specialized cells within the GI tract are capable of processing food proteins that survive the digestive process, and these include the microfold (M) cells, the intestinal epithelial cells, and dendritic cells. All 3 of these specialized cells of the GI tract play important roles in the presentation of antigen and in the development of oral tolerance.

M cells are specialized epithelial cells within the GI tract that are capable of transporting antigens from the lumen of the GI tract to antigen-presenting cells in the intestinal mucosa. M cells are located in the dome epithelium overlying Peyer patches and other gut-associated lymphoid tissue. Their microfolds and fenestrated membranes enable them to efficiently take up antigen and deliver them unprocessed to dendritic cells. Somewhat similarly, intestinal epithelial cells are capable of translocating antigen from the lumen of the GI tract to antigen-presenting cells by transcellular mechanisms. Both of these cells enable sampling of the proteins in the GI lumen and contribute to the development of oral tolerance within the intestinal milieu.

There are several different types of dendritic cells (DCs) within the GI mucosa, and it appears that these cells, involved in antigen processing and presentation, play a role in the development of tolerance. Antigen-sampling DCs that express the integrin CD11b and the chemokine receptor CX ₃ CR1 are capable of extending dendritic processes into the intestinal lumen. These cells are believed to collect antigen and then migrate into the mesenteric lymph nodes where they initiate activation and differentiation of effector T cells. A different class of DCs that expresses the integrin CD103 appear to be more preferentially involved in presenting antigen to naïve T cells and inducing their maturation into regulatory T cells (Tregs) for a more tolerogenic response.

In addition to suppressing autoimmune aberrant responses, Foxp3+ Treg cells are necessary for the development of oral tolerance. Foxp3-mutant mice, DEREG mice in which Foxp3 expression can be knocked out on treatment with diphtheria toxin, and IPEX (immunodysregulation polyendocrinopathy enteropathy X-linked syndrome) patients who have a mutation in the Foxp3 gene locus all have been shown to demonstrate the importance of Foxp3+ Tregs in the development of tolerance.

The ability for Tregs to induce a regulatory, tolerant immune response rather than an inflammatory response is facilitated in part by their elution of the inhibitory cytokines, transforming growth factor beta (TGF-ß) and interleukin-10. TGF-ß suppresses the downstream actions of T and B cells and promotes the production of secretory IgA. Meanwhile, IL-10 induces T-cell anergy, sustains Treg populations, and also plays a role in promoting B-cell class switching to produce secretory IgA.

Foxp3+ regulatory T cells generated in the mesenteric lymph nodes home to the mucosal surface in a manner that is dependent on retinoic-acid produced from dietary vitamin A. Although the exact mechanisms by which Tregs maintain oral tolerance in the luminal surface remain somewhat elusive, they are believed to play an important role in the maintenance (and perhaps development) of tolerance. Shreffler and colleagues demonstrated that milk-specific Foxp3+CD4+CD25 + Tregs were found in higher numbers in children who became tolerant to heated cow’s milk protein when compared with those who remained allergic to milk (and heated milk) and to healthy controls.

Sensitization

Our understanding of why some individuals develop allergic sensitization to food antigens while most individuals maintain immunologic tolerance is poor, but an increasing body of evidence suggests that the delicate balance between sensitization and tolerance may be affected at least in part by a few key determinants. These factors include the presence of adjuvants, dietary factors, doses and timing of antigen exposure, the properties of particular allergenic food components, the status of the microbiota, and the route of sensitization.

Like most humans, in the absence of an adjuvant, mice do not develop allergic sensitization to food antigens. Consequently, experimental adjuvants such as cholera toxin and staphylococcal enterotoxin B (SEB) have been required to induce hypersensitivity to food allergens.

Cholera toxin is a potent bacterial toxin that has been used as a mucosal adjuvant to break tolerance when coadministered with food antigens. Snider and colleagues were among the first to demonstrate that coadministration of cholera toxin with a food antigen induces antigen-specific IgE production and subsequent anaphylaxis as demonstrated by histamine release and drop in body temperature on reexposure. Additional studies by other investigators, including our laboratory, have demonstrated that similar protocols of coadministered cholera toxin and food antigens could be used to generate mouse models of milk-induced anaphylaxis and peanut-induced anaphylaxis. In total, this research has demonstrated that when cholera toxin is coadministered with a variety of food antigens, it induces a robust Th2-response, with antigen-specific IgE production, and anaphylaxis on reexposure. Despite the extensive use in these models, the precise mechanisms of action remain largely unclear but are beginning to be elucidated.

Blazquez and Berin demonstrated that cholera toxin alters the normally tolerogenic phenotype of CD103 + DCs, inducing their migration from the lamina propria to the mesenteric lymph nodes, where they upregulate the costimulatory surface molecules OX40L and Jagged-2 and induce Th2 cytokines, including IL-4 and IL-13, instead of undergoing their normal response of inducing Foxp3+Tregs. Blocking OX40L with neutralizing antibodies suppressed this Th2-skewed response while, in contrast, abolishment of Jagged-2 does not stop Th2-skewing, suggesting that OX40L plays a dominant role in this signaling cascade.

Obviously, in humans it is unlikely that cholera toxin plays a significant role in the development of food allergy, but analogous studies using SEB, a potent endotoxin known to be involved in human allergic disease, have suggested that SEB may play a role in the development of antigen sensitization. Yang and colleagues used coadministered SEB and OVA (Ovalbumin) to sensitize mice to ovalbumin and demonstrated that the combination of SEB plus OVA not only led to OVA sensitization with anaphylaxis on subsequent challenge, but also that this sensitization resulted in an upregulation of T-cell immunoglobulin and mucin domain molecule (TIM)-4 on intestinal dendritic cells. Subsequent studies demonstrated that blocking TIM-4 or its receptor TIM-1 prevented the sensitization and anaphylactic response. Our own studies demonstrated that SEB drove sensitization to OVA or whole peanut extract, that there was impairment in Treg functions, and that delivery of functionally capable Tregs to these mice was capable of reversing responses.

Our understanding of how these adjuvants lead to human food allergy remains incomplete, but the role of adjuvants seems to be an important aspect in the development of sensitization to foods. Potentially, factors that might affect the milieu in which antigen is processed within the GI tract, such as infection, intestinal injury, barrier function, mucosal integrity, and/or genetics, could have a similar influence on the immune system as adjuvants in the animal models and thereby may play an important role in breaking tolerance and skewing the immune response toward a Th2 phenotype.

Numerous studies have implicated the integrity of the skin and mucosal barrier in protecting against sensitization. The filaggrin gene is important in the development of an intact skin barrier; mutations in this gene are highly prevalent in patients with atopic dermatitis. Additional research has suggested that filaggrin mutations may lead to an increased risk of peanut allergy, presumably through deficiencies in skin defenses and/or gut defenses. Similarly, filaggrin mutations have been linked to eosinophilic esophagitis, a disease in which exposure to food proteins induces esophageal inflammation.

In addition to an intact mucosal barrier, it appears that there are factors within the mucosal epithelium that play an active role in maintaining tolerance and regulating immune responses. One such factor is thymic stromal lymphopoietin (TSLP). TSLP is constitutively expressed by the gut epithelial cells and is also expressed by stromal cells and basophils. TSLP activates OX40L on dendritic cells and generates a Th2 response that has been linked to the development of asthma and atopic dermatitis. In the GI tract, however, TSLP appears to play a more regulatory role, as Ziegler and Artis demonstrated that TSLP limits Th1-mediated and Th17-mediated inflammation in experimental models of helminth infections and colitis. TSLP’s regulatory role is complicated, however, as it is not required for sensitization or for oral tolerance. Further studies are needed to elucidate the role of TSLP in human food allergy.

Multiple studies have attempted to investigate the role of dietary factors on the development of food sensitization. Studies using data from the National Health and Nutrition Examination Survey (NHANES) have independently reported that vitamin D deficiency, folate levels, and increased obesity all may be associated with food-specific IgE levels. However, it should be noted, as has been pointed out by Keet and others, that food-specific IgE may not by itself be an appropriate proxy for the determination of clinical food allergy. Nonetheless, the concept that dietary factors may affect the immunologic milieu and subsequently affect the balance between sensitization and tolerance seems to be supported by these epidemiologic findings. Mechanistic support for the role of diet in the development of sensitization or tolerance has been shown in animal studies. In rats, a high-fat diet led to increased intestinal permeability, increased lipopolysaccharide in the serum, increased presence of inflammatory markers in the blood, and alterations in the gut microbiota, suggesting that changes in dietary components may alter overall immune regulation. However, further such studies evaluating the role of dietary factors in the development of food allergy, particularly as they relate to humans, are necessary.

Another factor that may contribute to the development of sensitization is the allergenic properties of the foods themselves. Although there is a tremendous amount and variety of food proteins in the world’s diet, only a small portion of them are responsible for most food allergies. These allergens are dominated by milk, soy, egg, wheat, fish, shellfish, peanut, and tree nuts. All of these foods possess a small molecular weight, an abundance of their relevant epitope(s), water solubility, glycosylation residues, and relative resistance to heat and digestion. It has been proposed that each of these characteristics contributes to the food allergens’ ability to survive the digestive process and to induce a Th2 response that results in IgE production and food allergy manifestations.

The timing of antigen exposure and dose of exposure also appear to be important factors in determining food antigen sensitization. Mouse models have demonstrated that high-dose exposure to an antigen early in life, even if given as a single dose, appears to induce T-cell anergy. Low-dose exposure, especially in repeated doses, has been shown to lead to the development of Tregs and thereby oral tolerance. In humans, numerous studies have suggested that delayed introduction of food antigens may lead to an increase in peanut allergy and wheat allergy. In a recent landmark study, Du Toit and colleagues demonstrated that early introduction of peanut to young infants who were at high risk of developing peanut allergy was protective against the development of peanut allergy. The mechanisms underlying the importance of timing and dose remain unclear, and further studies are needed in this area.

Another potentially important influence on the immune system’s response to a specific food antigen is the microbiota, or the microbial flora that comprise the intestinal environment. In mice, experiments using germ-free or antibiotic-treated mice have demonstrated an increased food-allergen sensitivity, elevated serum-specific IgE, and increased levels of circulating basophils. In humans, differences in the intestinal flora of allergic versus nonallergic children have been noted. Investigators have identified certain microbes that promote the development of Tregs and certain microbes that promote the skewing toward allergic sensitization. Although there remain a number of unknown details about the ways the microbiome affects the host’s susceptibility to food allergy, it is likely that the microbiome plays a significant role in the immunologic development of the GI tract and, more specifically, in shaping the immune response toward tolerance versus sensitization.

There have been a number of recent epidemiologic observations leading to the concept that the route of exposure to an antigen may be an important determinant of whether an individual becomes sensitized or tolerant. One such observation is that the incidence of food allergy continued to increase in the face of dietary measures and anticipatory guidance suggesting delayed introduction of allergenic foods. This finding suggests that sensitization may occur via routes other than ingestion. Another recent observation suggesting that sensitization may occur via the nonoral route arises from the discovery that most peanut-allergic children react on their first known ingestion of peanut. Importantly, peanut has been shown to be present in dust in the home and to associate with increased sensitization in children with skin barrier defects. These findings all suggest that the sensitization of some individuals might be occurring through nonoral exposures.

In mice, studies have found that sensitization may occur via an intragastric, sublingual, nasal, or cutaneous route. However, all of these routes of exposure required adjuvant to generate allergen alpha-lactalbumin (ALA)-specific IgE; in contrast, other studies have shown that epicutaneous sensitization can occur even in the absence of adjuvant. Interestingly, cutaneous exposure was associated with the highest level of ALA-specific IgE and the most severe anaphylactic responses in these animal models.

In humans, Fox and colleagues found that exposure to household peanut allergen was a risk factor for the development of peanut allergy in children, regardless of maternal ingestion during pregnancy or lactation. Their follow-up study demonstrated that peripheral blood mononuclear cells from peanut-allergic donors expressed the skin-homing integrin cutaneous lymphocyte antigen rather than the gut-homing integrin B7+, suggesting that the sensitization and subsequent lymphocyte imprinting occurred via the skin. Furthermore, in an intriguing observation that preempted the recently released landmark Learning Early About Peanut Allergies (LEAP) study, Du Toit and colleagues noted a 10-fold lower incidence of peanut allergy in Israeli children compared with their Jewish counterparts in London Hebrew schools. In Israel, children routinely consume a peanut-based snack, called Bamba, beginning in infancy, whereas in England (as in much of the rest of the Western world) peanut is often introduced later in childhood. Thus, the realization of Du Toit and colleagues suggested that not only may sensitization occur via a nonoral route (as appeared to be the case in the British children who were not consuming peanut), but also that early exposure via the oral route may be protective against the development of sensitization. The results of the LEAP study strongly suggest that early introduction of peanut by the oral route is highly protective against the development of sensitization and allergy. Additional studies will be needed to further understand the extent to which timing, dose, route of exposure, and property of the food themselves are involved in the development of sensitization or immunologic tolerance.