The intestine is an organ that shows the traces of evolutionary longevity, and indeed Cambrian period fossils from over 600 million years ago show a recognizable gastrointestinal tract.118 There is much current interest on the links between innate and adaptive immune responses, in particular pattern receptor molecules such as toll-like receptors and nod proteins that induce an immune response within innate cells, such as dendritic cells, that polarize subsequent T-cell responses.119,120 Evidence that oral tolerance cannot be established normally in germ-free mice suggests that the normal flora plays an important role in the generation of tolorogenic lymphocytes and the prevention of food allergies.121,122 The potential role for probiotics in prevention of food allergies in susceptible infants is thus likely to be based on the role of luminal bacteria in inducing a tolerant lymphocyte response.8,9 Transgenic mice whose only T cells responded to ovalbumin were in fact entirely tolerant of ovalbumin feeds, unless innate immune responses to the flora were blocked using cyclo-oxygenase-2 antagonists, when food-sensitive enteropathy was induced.123 Further study of dendritic cell populations within the intestine is likely to shed light on basic mechanisms of tolerance and sensitization. Genetic variation in receptors for bacterial products, such as toll-like receptors and nod proteins, is known to occur, and is likely to be related to allergic sensiti-zations,124 particularly as toll receptors are also expressed on mast cells.125
In addition to innate immune cells, the intestine also contains large numbers of primitive T and B cells, such as peritoneal B1 lymphocytes, y§ T cells, natural killer T cells and atypical CD8 cells with two a and no P chains.9 Little is known about the role of these primitive lymphocyte types in food allergies, although it is intriguing that yS-defi-cient mice have low mucosal IgA levels,126 while their numbers are increased within the mucosa in food allergies.30 There is also evidence that these types of T cell can react to lipid and glycolipid antigens, presented by non-classical MHC molecules such as CD1d, which are expressed by the enterocyte.127-129 It is thus notable that IgE responses to P-lactoglobulin in a rodent model of sensitization were substantially enhanced when the animals were sensitized to whole milk than to P-lactoglobulin alone.130
The role of extrathymically derived T cells, which mature within the gut epithelium, is also unknown but likely to be significant, particularly in infants born preterm. Given the numbers and situation of these cells, it is likely that further studies will unmask a relevant role in the induction of tolerance to dietary antigen. It is probably not coincidence that infants with a variety of immunodeficiencies have a high incidence of dietary sensitizations, chronic enteropathy and failure to thrive.131 It is likely that T-cell responses play a large role in determining tolerance or sensitization to dietary antigen. There are differences in racial susceptibility to sensitization to individual antigens, which do not relate to simple early life exposure.5
One cell type that is likely to play a more significant role in food allergies than was previously recognized is the small-intestinal enterocyte, which separates the immune system from both dietary antigens and bacteria. Their role in antigen presentation and production of an array of cytokines has recently been recognized.32,128,129,132
An important role in induction of tolerance has thus been suggested. Recent data suggest that intestinal epithelium may directly induce a regulatory phenotype in CD8 cells.133 Increased paracel-lular permeability, allowing transfer of antigen to the immune system without epithelial processing and presentation, may underlie the phenomenon of sensitization to food antigens during conditions such as rotavirus gastroenteritis.32,97 Formal confirmation of the sensitizing role of excess paracellu-lar permeability was provided in a transgenic mouse model where an intercellular adhesion molecule called cadherin was mutated, inducing severe enteropathy.134
While non-IgE-mediated responses to dietary antigen may cause chronic symptomatology, IgE-mediated mechanisms account for the majority of immediate hypersensitive reactions to foods. Transient IgE responses to foods are seen in normal children, so this is unlikely to be clinically relevant.135 By contrast, high-level IgE responses are usually pathological and may be important in severe food allergies and anaphylaxis.
Production of IgE is favored by dominance of Th2 responses, particularly due to IL-4 and IL-13 secretion.136 The receptors for IL-4 and IL-13 share a common a chain (IL-4Ra), mutations in which increased signaling is associated with increased atopy.137,138 Intracellular signaling downstream of this receptor is mediated through Stat-6 (signal transducer and activator of transcription-6), and blockade of either IL-4Ra or Stat-6 appears promising as a therapeutic target for IgE-mediated allergic reactions.139,140
Lineage commitment of B cells towards IgE is favored by the presence of IL-4 and IL-13, and inhibited by Th1-associated cytokines.136 This may potentially occur within the Peyer's patches in circumstances of Th2-skewed local responses, and lead to generation of mucosal IgE-producing plasma cells without necessarily affecting the commitment of circulating B-cell populations. Mucosally produced IgE may also be transported into the lumen of the gut or airway by a mechanism distinct from secretory component-mediated IgA transport.141,142 It is thus possible that a compartmentalized response may occur within the intestine, in which mucosal IgE responses may be elicited by dietary antigen, even if cutaneous IgE responses do not occur, leading to negative skin prick tests, and in the absence of circulating specific IgE.
There are undoubted links between intestinal food allergic responses and the infiltration of both eosinophils and mast cells, which produce a variety of vasoactive and neuroactive mediators. Both cell types have been particularly implicated in dysmotility responses, and it is likely that these products may directly affect the function of enteric nerves.143 During the food allergic response, both mast cell tryptase and cationic protein (ECP) are released into the lumen and may be detected in stools.144-146 However, although mast cells and eosinophils produce a similar spectrum of mediators, their responses show a different time-course. Mast cells induce rapid responses through immediate degranulation of mediators stored in intracel-lular granules, whereas eosinophil responses are often delayed for several hours, as they are recruited from the peripheral circulation into tissues.
Two molecules are very clearly implicated in mucosal eosinophilia - the chemokine eotaxin and the cytokine IL-5. Important studies by Rothenberg et al, using targeted gene deletion in mice, have clarified the relative contributions of both mediators. Mice were sensitized to ovalbumin, and then challenged with oral administration of ovalbumin-coated beads.147 Wild-type mice mounted an allergen-specific Th2 response, showing mucosal eosinophilia with increased circulating IgE and IgG1, while eotaxin-deficient mice had preserved systemic IgE and IgG1 responses but did not recruit eosinophils to the mucosa. By contrast, IL5-deficient mice had reduced circulating eosinophils.147 In additional studies of gastric eosinophil recruitment, it was possible to confirm that antigen-coated beads induced delayed gastric emptying, and again eotaxin-deficient mice were protected.148 Further studies from this group have identified the esophagus as an apparent target for eosinophil recruitment, either as a consequence of inhalation of aeroallergens such as aspergillus149 or in circumstances of excess systemic IL-5 expression.150 This targeting may presumably have some as yet unclear evolutionary basis, but these findings have potentially important clinical implications. First, there may be more than one factor inducing esophageal eosinophilia, and there may be only a partial response to antigen exclusion if the child is also responding to inhaled aspergillus. Second, an intercurrent systemic viral illness in an allergic child with a Th2-deviated immune response may promote esophageal eosinophilia in an antigen non-specific manner. The increased frequency of viral infections in food-allergic children, probably due to low immunoglobulins, CD8 and natural killer cells,14 may make this a clinically difficult scenario, as appropriate food exclusions may give an apparently poor clinical response.
In adults, IL-5 mRNA is increased within the small-bowel mucosa of food-allergic patients, but not in atopic or non-allergic controls.151 T cells from food antigen sensitized children produce IL-5 on food challenge, whereas those from tolerant children do not.93,94 Thus, there is evidence for a final common pathway in the mucosal allergic response to dietary antigen, which is dependent on up-regulation of IL-5 production and expression of the chemokine eotaxin.
The phenomenon of oral tolerance lies at the heart of food allergy, which by its nature implies a breakdown or failure of establishment of oral tolerance mechanisms. There has been much recent progress in the understanding of oral tolerance mechanisms. The dose of ingested antigen appears to be particularly important in determining how tolerance is established.152 The bulk of dietary antigen is absorbed by enterocytes for nutritional purposes, and this antigen is presented by the epithelium in such a way that lymphocyte reactivity is suppressed and the lymphocytes become anergic.128,129 The absent expression of co-stimulatory molecules by enterocytes and production of suppressor cytokines may both play a role, although little is known about these processes in human infancy and childhood. Food antigens have also been shown to induce apoptosis of antigen-specific lymphocytes in the Peyer's patches of mice, but this has not yet been shown in humans.153 More recent data suggest that a more complex state is induced in tolerogenic lymphocytes by food administration, in which pro-apop-totic and anti-apoptotic factors are simultaneously up-regulated while T-cell receptor signaling molecules are down-regulated.154
In contrast, tolerance to low doses of antigen requires uptake by the antigen-sampling M cells that overlie Peyer's patches. This form of tolerance requires an active generation of suppressor lymphocytes within the Peyer's patches, and in particular Th3 cells that produce the anti-inflammatory cytokine TGF-p.8,155,156 Other regulatory T-cell populations that are likely to be very important in preventing food allergies include T regulator-1 (Tr1) cells, which produce IL-10, and CD4 + CD25+ cells.9,129,156 The transcription factor Foxp3 appears to be central in the commitment of naive T cells towards the regulatory pathway in mice.157 Similar relevance in humans is suggested by the development of a multifocal inflammatory condition (IPEX syndrome) in infants with mutations in Foxp3.158 Thus, Foxp3 may be a molecule that is of critical importance in maintaining oral and systemic tolerance. There are currently no clear data to determine which regulatory cell type is most relevant in prevention of childhood food allergy. There is certainly evidence that a multiply exposed population of elderly circulating CD4 + CD25+ cells inhibits milk responses in adult humans.159 However, which would be unlikely to be the case in early life, when most T cells are initially of the naive phenotype? Analysis of circulating CD4 + CD25+ in adults shows that the majority express the skin-homing marker CLA, but not the gut-homing P7 integrin, while cord blood CD25 + CD4+ cells express neither CLA nor P7 integrin.160 Evidence of functional immaturity of the neonatal CD4 + CD25+ cell population, at least in mice, is provided by data showing that adult CD4 + CD25+ cells prevent an autoimmune response to the autoantigen myelin oligodendro-cyte protein, whereas cord blood cells do not.161 This is likely to be functionally important in responses to ingested antigen, as neonatal animals show impaired low-dose oral tolerance and may even paradoxically sensitize.162
There are now extensive data to suggest that oral tolerance is not innate, and that the mechanisms are not present at birth but develop postnatally. The expression of a specific array of toll receptors on CD4 + CD25+ regulatory T cells, and increase of their suppressor functions by bacterial lipopolysaccharide,163 suggest that the early infectious exposures of the young infant are indeed likely to be important in the generation of oral tolerance and the prevention of allergy. The current data on human infants point to a particular role for TGF-P rather than IL-10 in the prevention of infant allergy. Study of spontaneous and cow's milk-stimulated cytokine production of cord blood mononuclear cells identified a reduced TGF-P but not IL-10 response in children of allergic mothers.164 The dominant cytokine abnormality in the mucosa of food-allergic children does not appear to be diminished Th1 or excess Th2 responses, but reduced numbers of TGF-P secreting Th3 cells.34 Other studies have confirmed that Th1 responses are normal or even increased within the mucosa in childhood food allergy,31,33,34 but that TGF-P responses are impaired: expression of TGF-P1 and its receptor are diminished within the mucosa in food-allergic enterocolitis,165 while milk-reactive T-cell clones from milk-allergic children show a Th2-deviated response but with minimal TGF-P production.166 The factors involved in early life generation of TGF-p1-produc-ing Th3 cells are thus likely to be of great importance in determining whether food-allergic sensiti-zation occurs. Infectious exposures are certainly one such factor, and it is notable that the density of TGF-P-producing cells within the duodenal mucosa of infants in rural Gambia was an order of magnitude higher than in healthy UK infants.167 It is likely that early-life immunomodulation of regulatory responses, rather than alteration of dietary exposures, will prove the way ahead in childhood food allergies.
Was this article helpful?