Food allergy (FA) has several distinct clinical and immunologic entities. The majority of young children have a classic type I, IgE-mediated immediate immune response. In delayed-type food allergy, intestinal symptoms appear over several hours or even days after exposure to the allergen, and the reaction is non-IgE mediated (1). Although cytokine imbalance has been proposed to be important in the pathogenesis of delayed-type FA (2,3), only one study has addressed this in the target organ, the gut (4). In that study, Hauer et al, using enzyme-linked immunoabsorbant spot technique, showed an elevation in IFN-γ and, to a lesser extent, IL-4 secretion in duodenal biopsies of infants with FA. Cells secreting IFN-γ were ten times as abundant as cells secreting IL-4, indicating a T helper cell type 1 (Th1) dominance. However, to our knowledge, no previous work seems to have characterized the cytokine profiles in the intestine of patients with FA, by evaluating the expression of both messenger RNA and proteins.
Lymphocyte recruitment at the sites of allergic reaction is a complex event involving the expression of surface adhesion molecules on circulating cells, and interaction with their counter-receptors on vascular endothelial cells (5). The interaction of the α4β7 integrin on lymphocytes has been shown to be responsible for their homing to mucosal sites in the gut (6,7). The integrin lymphocyte function-associated antigen-1 (LFA-1 or CD11a/CD18), which belongs to the β2 family of integrin receptors, and its complementary molecule intercellular adhesion molecule-1 (ICAM-1) play an important role in the firm adhesion and migration of activated lymphocytes in allergic reactions (8).
In this study we evaluated the mRNA expression of IL-4 and IFN-γ by radioactive in situ hybridization in the duodenal mucosa of children with untreated and treated food allergy. We also investigated the protein expression of IL-4 and IFN-γ and the expression of adhesion molecules (LFA-1, ICAM-1, α4β7) by immunohistochemistry in the same samples. Further, we quantified lymphocyte subgroups, and the dividing epithelial cells in the crypts, and evaluated the expression of MHC class II antigens as markers of immune activation.
MATERIALS AND METHODS
Patients and controls
The study included 7 children (mean age 7.3 years, range 2-13) with untreated FA (uFA) and 7 children (mean age 8.1, range 1-14) with treated FA (tFA), still having some clinical symptoms or growth retardation as an indication for gastroduodenoscopy, and 5 controls (mean age 11.4, range 4-16). All patients had been referred to the Oulu University Hospital for a pediatric gastroenterological consultation because of recurrent abdominal pains, prolonged diarrhea and/or growth retardation. Cases with any suspicion of FA underwent an open 6-week elimination/challenge test as described below. The control patients were studied for diagnostic purposes and were all subsequently diagnosed as having normal endoscopic and histologic results with no history of allergy. Celiac disease was excluded by serological testing and the finding of normal villous structure in the duodenal specimens. Lactose intolerance was also ruled out. The morphology of all mucosal specimens of patients with uFA and tFA, and of controls was normal at routine histologic examination. Clinical data of the patients with FA are summarized in Table 1.
In addition, 5 patients with celiac disease (mean age 7.0, range 2-15) were included, to serve exclusively as positive controls for in situ hybridization. Biopsy specimens from patients with celiac disease showed severe partial or subtotal villous atrophy with crypt hyperplasia and the patients had positive serum antiendomysial antibodies.
Diagnosis of food allergy
Diagnosis of FA was based on the disappearance of symptoms on an elimination diet (2 weeks) and their reappearance in an open challenge test as described in detail in our previous study (9). Briefly, oral challenges were performed with milk and cereals (gluten-containing cereals and oats, separately); each period continued for 2 weeks if symptoms did not manifest earlier. Abdominal pain, diarrhea or loose mucous stools, or exacerbation of dermatitis were each considered a positive result. Of the 7 patients with uFA, 4 reacted to milk, one to cereals, and 2 to both milk and cereals. In the tFA group, 4 reacted to milk, 2 to cereals, and 1 to both milk and cereals. In all cases symptoms appeared at earliest after 24 hours from the beginning of the challenge. RAST test was performed in 5 of the 7 patients with uFA and in 2 out of the 7 patients with tFA. In addition, skin test was performed in one of the patients in the tFA group. The results can be seen in Table 1.
Endoscopic examination and samples
Upper intestinal endoscopy was performed with an Olympus GIF-XQ140 under general anesthesia. The assessment of lymphoid nodularity on the mucosa of the duodenal bulb was based on an endoscopic evaluation after the bulb was filled with air. Biopsies for routine histology and for immunohistochemistry and in situ hybridization were drawn from the bulb of the duodenum. The specimens were embedded in optimal cutting temperature (OCT) compound (Miles Laboratories, Elkhart, IN), snap frozen in liquid nitrogen, and stored at -70°C until used. Frozen tissue samples were cut into 8-μm sections for immunohistochemistry and 10-μm sections for in situ hybridization. The sections were coded and evaluated without knowledge of the specimen.
Immunohistochemistry and microscopical evaluation
Single immunostaining of mononuclear cells for CD3, CD4, CD8, γδ- and αβ-TCR, CD25 (IL-2R), CD11a (LFA-1), and α4β7 integrin; mucosal staining of the HLA class II (HLA-DR) and ICAM-1; and staining of proliferative epithelial cells with Ki-67 were all performed on native cryostat sections by the avidin-biotin immunoperoxidase system as described earlier (10,11). For immunostaining of IL-4 and IFN-γ, permeabilization was performed by incubation of the slides in 0.1% PBSTween20 for 10 minutes in RT on a shaking platform before the slides were incubated with monoclonal antibodies diluted in 1% normal serum in 0.1% PBS-Tween20 for one hour at +37°C. The positively stained cells were counted under a light microscope with a calibrated graticule at x1000 magnification as described earlier in detail (11). In the same specimen, the positive cells in at least 30 fields either along the epithelium or comprising lamina propria were counted, and cell densities expressed as cells/mm or cells/mm2, respectively. Ki-67 positive cells were calculated as percentages of the crypt epithelial cells, with at least 200 crypt cells counted in each specimen. The proportions of HLA-DR positive epithelial cells and ICAM-1 positive lamina propria cells were measured by point counting, with a grid with 100 points. The monoclonal primary antibodies are listed in Table 2.
Radioactive RNA in situ hybridization
Samples were the same as used for immunohistochemistry, and serial sections from the biopsies were processed for in situ hybridization. Because of the small size of the specimens, we lacked adequate material from patient number 8 and from one of the controls for in situ hybridization.
Frozen tissue samples were cut into 10-μm sections and fixed in 4% paraformaldehyde in PBS. The sections were subjected to in situ hybridization for human IL-4 and IFN-γ riboprobes obtained from the cDNAs as described earlier (12). Tissue sections were incubated with 1.2 × 106 CPM of [33P]-labeled (1000-3000 Ci/mmol; Amersham, Life Technologies, Arlington Heights, IL) antisense or sense riboprobe in a total volume of 80 μl following the in situ protocol described earlier (13).
Microscopic evaluation of RNA in situ hybridization
Positive messenger RNA (mRNA) expression for IL-4 and IFN-γ was analyzed as described previously (12). Briefly, positive mRNA expression was counted in 2 representative areas in the lamina propria, as well as in the background staining adjacent to the tissue in high-power fields (magnification ×400). The relative intensity of the positive mRNA signal was determined by dividing the positivity in the tissue by the positivity of the background. The lamina propria was assessed from the area immediately below the surface epithelium, excluding the epithelia and areas with lymphatic aggregates. For the specific probes, a minimum of 2 sections was prepared for each patient investigated.
Cell densities and the percentages of positive staining within the 4 study groups were compared by non-parametric tests (Mann-Whitney U test) because of the non-linear distribution of the parameters. A P value <0.05 was considered significant.
For an excessive biopsy sample, after an oral explanation of the study plan, a written parental consent with a signature was obtained from the parents of all children. The protocol was approved by the ethics committee at Oulu University Hospital.
The density of δTCR+ intraepithelial lymphocytes was significantly higher in children with uFA than in controls (P = 0.010). The density of these cells in patients with tFA was similar to that of controls (Table 3). The densities of CD3-, CD4-, CD8-, and αβ-TCR-positive cells in the intraepithelial compartment were similar in patients and in controls (only data for CD3 shown, Table 3).
Expression of adhesion molecules in the epithelium
The density of LFA-1+ intraepithelial cells in patients and in controls was similar. The density of α4β7 +cells in the epithelium of patients with uFA tended to be greater than in those with tFA or controls, but did not reach significance (Table 3).
Lymphocytes in the lamina propria
Controls tended to have a higher density of CD3+ cells in the lamina propria than did the 2 FA groups (Table 4). The densities of CD4-, CD8- and αβ-TCR-positive cells in the lamina propria in patients and in controls were similar (data not shown). The densities of CD25+ cells in the 3 study groups were similar (Table 4).
Expression of adhesion molecules in the lamina propria
The lamina propria of uFA patients contained more cells bearing LFA-1 integrin than did those of tFA or controls, although the differences were non-significant (P=0.128 and P=0.149, respectively). The density of α4β7 + cells in the lamina propria of controls tended to be higher than that of children with FA, although these differences did not attain significance (tFA P = 0.073 and uFA P = 0.106) (Table 4). No significant difference appeared in the expression of ICAM-1 on endothelial cells of capillaries or on mononuclear cells in the mucosal lamina propria of the different study groups. The surface enterocytes and intraepithelial lymphocytes showed complete absence of ICAM-1 expression (Table 5).
Cytokines in the lamina propria
IFN-γ+ cells were confined to the lamina propria, with a higher density of IFN-γ+ cells in the lamina propria of children with uFA than with tFA or controls (P = 0.053 and P = 0.018, respectively) (Fig. 1). Only scattered IFN-γ+ and IL-4+ cells appeared in the epithelium (0-1 cells/mm; data not shown). The numbers of IL-4-positive cells in the lamina propria were similar in all groups (Table 4).
Staining with Ki-67
The percentage of proliferative cells in the crypts (Ki-67) of specimens from uFA was significantly higher than in patients with tFA (P=0.026) and tended to be higher than in controls, indicating higher turnover rates of epithelial cells in the untreated group (Table 5).
Staining with HLA-DR antibody
Significantly stronger positive staining of crypt cells with HLA-DR was present in the uFA than in controls (P = 0.048). Surface epithelium stained with HLA-DR antibodies was significantly higher in tFA than in controls (P = 0.048) (Table 5). We found increased HLA-DR staining on lamina propria cells both in uFA and tFA patients; however, the difference was significant only when tFA was compared to control subjects (P=0.030) (Table 4).
Radioactive RNA in situ hybridization
IL-4 mRNA expression
The duodenal mucosa of all patients studied contained cells expressing mRNA for IL-4. In all specimens, intestinal mononuclear cells scattered in the lamina propria were found to express IL-4 mRNA, some positive signal also being found extracellularly in a diffuse fashion. The expression of IL-4 mRNA was significantly higher in celiac patients than in the other study groups (uFA, tFA, controls, P = 0.006, P = 0.010 and P = 0.029, respectively). The expression of IL-4 in FA patients did not differ from that of controls (Fig. 2). All sections hybridized with the IL-4 sense probe showed only background signals.
IFN-γ mRNA expression
In situ hybridization with the antisense probe for IFN-γ showed that positive cells were found mainly in the superficial lamina propria. The surface epithelium and intraepithelial lymphocytes in specimens of FA patients and controls showed very little expression of IFN-γ. The expression of IFN-γ mRNA in the lamina propria was significantly higher in celiac specimens than in tFA or controls (P=0.017 and P=0.016, respectively), but not in uFA (Fig. 2). As for IL-4, the IFN-γ sense probe expressed only background signals.
This study shows immune activation in the small intestine of children with delayed-type food allergy. We found significantly higher density of IFN-γ+ cells in the lamina propria of children with untreated FA than in controls. The expression of IFN-γ mRNA detected by in situ hybridization in the groups showed the same tendency as protein expression. Further, despite the normal villous structure, the untreated FA patients exhibited a higher crypt proliferation rate and HLA-DR crypt staining, and an increment in the density of γδ-TCR+ intraepithelial lymphocytes. Most of these differences tended to be normalized in patients on an exclusion diet (tFA).
Patchy villous atrophy with cellular infiltration is considered one of the characteristics of food protein-induced enteropathy (1). Our patients with FA were significantly different from those of Hauer et al. (4) and other reports, being older and presenting mainly with digestive symptoms. Still, differences in the expressions of several markers were evident. Our group has recently characterized delayed type of food allergy extending beyond infancy (14). In this entity only a minority of patients are atopic and have atopic dermatitis. Their intestines do not show eosinophilic or mononuclear infiltration, or even minor villous alterations, features typical for small children with FA. A prominent finding in these patients has been lymphonodular hyperplasia, described also in all but one of the patients in this study. However, we can not exclude the role of IgE-mediated immune response in these patients, although the delayed onset of symptoms at challenge strongly speaks against an IgE mediation.
Our previous work showed increased adhesion molecules and HLA-DR markers on structurally normal duodenal biopsies of adult patients with FA (15). Such subtle changes are also seen in the intestine of patients with an early phase of celiac disease: during its latency before the evolution of villous atrophy, an elevated number of intraepithelial lymphocytes, especially γδ-TCR+ cells, and signs of activated mucosal cell-mediated immunity, such as expression of CD25 and B7, and upregulation of MHC II molecules and of adhesion molecules may be revealed by immunohistochemical methods (16,17). Further, in our previous work a high density of γδ-TCR+ intraepithelial lymphocytes seemed to be associated with the presence of increased densities of IFN-γ+ and TNF-α+ cells in the lamina propria of potential celiac jejunal specimens (12). However, our and other previous results indicate that the mucosal changes in FA arise from another pathogenic mechanism than in CD, and FA can not be considered a precursor to CD (14).
Both LFA-1 and ICAM-1 are suggested to be important in the pathomechanism of the early phase of celiac disease (18). In experimental murine food-hypersensitive crypt hyperplasia, after antigen challenge, increased numbers of intraepithelial lymphocytes, and enhanced LFA-1 expression on the lamina propria T cells are seen in gut mucosa (19). We found no significant difference in the expression of ICAM-1 in the lamina propria of patients with FA or controls, which is consistent with previously published results in human beings (20). The surface enterocytes and intraepithelial lymphocytes showed a complete absence of ICAM-1 expression, indicating that antigen presentation and induction of co-stimulatory signals occur mainly in the lamina propria.
Our previous work showed integrin α4β7, the guthoming receptor known to bind to mucosal addressin cell adhesion molecule-1 in the intestinal mucosa (5), to be upregulated in the duodenal lamina propria of adult patients with FA (15). These patients exhibited a markedly elevated density of α4β7 + cells in the epithelium, and increased numbers of LFA-1+ cells in the lamina propria. In the present study, we found a similar tendency concerning α4β7 and LFA-1. In children with untreated FA, the former was higher in the intraepithelial compartment and the latter in the lamina propria. Nonetheless, it should be noted that these differences failed to reach statistical significance. The route of immunization by an antigen probably determines the type of homing receptor memory that the T cell is expressing. The α4β7 receptor on T cells suggests that the antigen encounter has taken place in the intestine (21,22). It has recently also been shown that strongly α4β7-positive CD4+ T cells were more likely to produce IFN-γ than were α4β7-negative CD4+ T cells, suggesting a connection between these homing receptor-bearing cells and the Th1 response (23).
The increased density of IFN-γ+ cells in the lamina propria of children beyond infancy with untreated FA suggests a Th1 response leading to cell-mediated mechanisms. IFN-γ is known to be produced mainly by activated CD4+ Th1 cells in the lamina propria, inducing epithelial crypt cell hyperplasia and cell activation, as well as local proliferation of intraepithelial lymphocytes. In addition, IFN-γ induces the expression of MCH-class II antigens on the epithelial cells (24). We found both of these associations in the present study. The expression of IFN-γ mRNA detected by radioactive in situ hybridization in the lamina propria of patients with uFA was higher than in tFA and controls, as suggested by comparing the results of these study groups with the results of patients with celiac disease. However, while by immunohistochemistry, the differences between patients with uFA and tFA or controls were not significant. This difference between IFN-γ mRNA and protein expression may be caused by the IFN-γ cells not being highly activated, thus expressing less mRNA. Similar finding concerning IFN-γ cells was documented in a previous study investigating the cytokine profiles of freshly isolated mononuclear cells from Peyer's patches of healthy pediatric patients by ELISPOT and quantitative RT-PCR (25).
IL-4 is predominantly produced by T lymphocytes and seems to have effects differing from those of IFN-γ on a wide variety of immune functions (26). In cow's milk enteropathy, Hauer et al detected higher numbers of IL-4 and IFN-γ secreting cells from the duodenal mucosa of infants than from controls (4). However, IFN-γ secreting cells were about tenfold more abundant than IL-4 secreting cells, suggesting an exaggerated Th1 response. We found no difference concerning IL-4 mRNA- and IL-4 protein-positive cells in the FA and control study groups. However, both the type of patients and methods used differed between these studies.
The cytokine-positive cells, both IFN-γ and IL-4, were distributed in the lamina propria in relatively high numbers also in controls as described earlier (4,12,27). The presence of significant numbers of IFN-γ and IL-4 containing cells in the normal small intestinal mucosa may indicate a constant state of activation. It has been suggested that the increased production of IFN-γ seen in the mucosa in Crohn disease (28) and celiac disease (29) represents an exaggeration of the normal physiologic response. We found only scattered IFN-γ+ and IL-4+ cells in the surface epithelium, which is inconsistent with previous findings where a larger proportion of epithelial lymphocytes spontaneously secreted IFN-γ and IL-4 than lamina propria lymphocytes did (27). However, this discrepancy may be caused by the different techniques used. We did not analyze a suspension of isolated intraepithelial lymphocytes, and therefor the proportion of IELs secreting IFN-γ or IL-4 could not be determined.
The higher expression of γδ-TCR+ intraepithelial lymphocytes of children with untreated FA is in agreement with earlier observations (9,30,31). This increment was associated with the active and untreated phase of FA, as the subjects currently on a diet for FA exhibited low levels of γδ-T cells, comparable to those of controls. In contrast, the high density of γδ-T cells observed in celiac disease is much more pronounced and is observable for a prolonged period (10,30). The exact function of these γδ-TCR+ cells is, as yet, unclear. Intraepithelial γδ-TCR+ cells can produce a multiplicity of cytokines, including IL-2, IFN-γ, TNF-α, IL-4, IL-10, and TGF-β (32). These may regulate the growth and differentiation of intestinal epithelial cells (33). In addition, their role in gluten or food hypersensitivity may be related to maintaining or abrogating oral tolerance; γδ-TCR+ cells from a non-tolerant animal slow down the development of tolerance (34). Abrogation of tolerance leads to enhanced immune responses to the antigens and may be associated with intestinal pathology (35).
In conclusion, our study showed despite normal villous structure increased expression of IFN-γ-positive cells in the lamina propria of children with delayed-type FA, indicating a Th1 response. This finding suggests that cytokine imbalance may play an important role in the pathogenesis of delayed-type FA, inducing crypt cell turnover and activation of epithelial cells.
We are grateful for the skillful technical assistance of Ms. Sirkku Kristiansen and thank Prof. Markku Heikinheimo and Ilkka Ketola, M.D., Hospital for Children and Adolescents, University of Helsinki, for their expert help with radioactive RNA in situ hybridization.
1. Sampson HA. Food allergy. Part 1: immunopathogenesis and clinical disorders. J Allergy Clin Immunol
2. Hill DJ, Ball G, Hosking CS, et al. Gamma-interferon production in cow milk allergy. Allergy
3. Sütas Y, Hurme M, Isolauri E. Oral cow milk challenge abolishes antigen-specific interferon-gamma production in the peripheral blood of children with atopic dermatitis and cow milk allergy. Clin Exp Allergy
4. Hauer AC, Breese EJ, Walker-Smith JA, et al. The frequency of cells secreting interferon-gamma and interleukin-4, -5, and -10 in the blood and duodenal mucosa of children with cow's milk hypersensitivity. Pediatr Res
5. Picker LJ. Control of lymphocyte homing. Curr Opin Immunol
6. Berlin C, Berg EL, Briskin MJ, et al. Alpha 4 beta 7 integrin mediates lymphocyte binding to the mucosal vascular addressin MAdCAM-1. Cell
7. Briskin M, Winsor-Hines D, Shyjan A, et al. Human mucosal addressin cell adhesion molecule-1 is preferentially expressed in intestinal tract and associated lymphoid tissue. Am J Pathol
8. Wegner CD, Gundel RH, Reilly P, et al. Intercellular adhesion molecule-1 (ICAM-1) in the pathogenesis of asthma. Science
9. Kokkonen J, Holm K, Karttunen TJ, et al. Children with untreated food allergy express a relative increment in the density of duodenal gammadelta+ T cells. Scand J Gastroenterol
10. Savilahti E, Arato A, Verkasalo M. Intestinal gamma/delta receptor-bearing T lymphocytes in celiac disease and inflammatory bowel diseases in children. Constant increase in celiac disease. Pediatr Res
11. Savilahti E, Örmälä T, Saukkonen T, et al. Jejuna of patients with insulin-dependent diabetes mellitus (IDDM) show signs of immune activation. Clin Exp Immunol
12. Westerholm-Ormio M, Garioch J, Ketola I, et al. Inflammatory cytokines
in small intestinal mucosa of patients with potential coeliac disease. Clin Exp Immunol
13. Wilkinson D. Whole mount in situ hybridization of vertebra embryos. In: D Wilkinson, ed. In situ hybridization- a practical approach
. Oxford: IRL Press; 1992:75-83.
14. Kokkonen J, Haapalahti M, Laurila K, et al. Cow's milk proteinsensitive enteropathy at school age. J Pediatr
15. Veres G, Helin T, Arato A, et al. Increased expression of intercellular adhesion molecule-1 and mucosal adhesion molecule alpha4beta7 integrin in small intestinal mucosa of adult patients with food allergy. Clin Immunol
16. Mäki M, Holm K, Collin P, et al. Increase in gamma/delta T cell receptor bearing lymphocytes in normal small bowel mucosa in latent coeliac disease. Gut
17. Holm K, Savilahti E, Koskimies S, et al. Immunohistochemical changes in the jejunum in first degree relatives of patients with coeliac disease and the coeliac disease marker DQ genes. HLA class II antigen expression, interleukin-2 receptor positive cells and dividing crypt cells. Gut
18. Maiuri L, Picarelli A, Boirivant M, et al. Definition of the initial immunologic modifications upon in vitro gliadin challenge in the small intestine of celiac patients. Gastroenterology
19. Ohtsuka Y, Yamashiro Y, Maeda M, et al. Food antigen activates intraepithelial and lamina propria lymphocytes in food-sensitive enteropathy in mice. Pediatr Res
20. Chung HL, Hwang JB, Kwon YD, et al. Deposition of eosinophilgranule major basic protein and expression of intercellular adhesion molecule-1 and vascular cell adhesion molecule-1 in the mucosa of the small intestine in infants with cow's milk-sensitive enteropathy. J Allergy Clin Immunol
21. Eigenmann PA, Tropia L, Hauser C. The mucosal adhesion receptor alpha4beta7 integrin is selectively increased in lymphocytes stimulated with beta-lactoglobulin in children allergic to cow's milk. J Allergy Clin Immunol
22. Kantele A, Zivny J, Häkkinen M, et al. Differential homing commitments of antigen-specific T cells after oral or parenteral immunization in humans. J Immunol
23. Abramson O, Qiu S, Erle DJ. Preferential production of interferon-gamma by CD4+ T cells expressing the homing receptor integrin alpha4/beta7. Immunology
24. Terpend K, Boisgerault F, Blaton MA, et al. Protein transport and processing by human HT29-19A intestinal cells: effect of interferon gamma. Gut
25. Hauer AC, Bajaj-Elliott M, Williams CB, et al. An analysis of interferon gamma, IL-4, IL-5 and IL-10 production by ELISPOT and quantitative reverse transcriptase-PCR in human Peyer's patches. Cytokine
26. Mosmann TR, Coffman RL. TH1 and TH2 cells: different patterns of lymphokine secretion lead to different functional properties. Annu Rev Immunol
27. Carol M, Lambrechts A, Van Gossum A, et al. Spontaneous secretion of interferon gamma and interleukin 4 by human intraepithelial and lamina propria gut lymphocytes. Gut
28. Breese E, Braegger CP, Corrigan CJ, et al. Interleukin-2- and interferon-gamma-secreting T cells in normal and diseased human intestinal mucosa. Immunology
29. Nilsen EM, Jahnsen FL, Lundin KE, et al. Gluten induces an intestinal cytokine response strongly dominated by interferon gamma in patients with celiac disease. Gastroenterology
30. Spencer J, Isaacson PG, MacDonald TT, et al. Gamma/delta T cells and the diagnosis of coeliac disease. Clin Exp Immunol
31. Savilahti E. Food-induced malabsorption syndromes. J Pediatr Gastroenterol Nutr
32. Haas W, Pereira P, Tonegawa S. Gamma/delta cells. Annu Rev Immunol
33. Boismenu R, Havran WL. Modulation of epithelial cell growth by intraepithelial gamma delta T cells. Science
34. Fujihashi K, Taguchi T, Aicher WK, et al. Immunoregulatory functions for murine intraepithelial lymphocytes: gamma/delta T cell receptor-positive (TCR+) T cells abrogate oral tolerance, while alpha/beta TCR+ T cells provide B cell help. J Exp Med
35. Barbeau WE. Interactions between dietary proteins and the human system: implications for oral tolerance and food-related diseases. Adv Exp Med Biol