Advertisement

Antigen-specific T-cell responses in patients with non–IgE-mediated gastrointestinal food allergy are predominantly skewed to TH2

Published:October 18, 2012DOI:https://doi.org/10.1016/j.jaci.2012.09.005
      To the Editor:
      IgE-mediated allergy is triggered by cross-linking of antigen-specific IgE antibodies on the cell surfaces of mast cells and basophils, followed by local accumulation and activation of inflammatory cells, including eosinophils and TH2 cells. TH2 cells produce such cytokines as IL-4, IL-5, and IL-13, which promote IgE production and eosinophilopoiesis and play central roles in the development of chronic allergic inflammation. On the other hand, non–IgE-mediated allergies, such as hypersensitivity pneumonitis, are considered mediated by cellular immunity, which has not been thought to involve antigen-specific TH2 cells because IgE antibody would be detected if TH2 cells were activated. Non–IgE-mediated gastrointestinal food allergies include food protein–induced enterocolitis syndrome (FPIES), food protein–induced proctocolitis, and food protein–induced enteropathy. The precise underlying mechanisms are almost unknown, except for a fundamental role of TNF-α,
      • Chung H.L.
      • Hwang J.B.
      • Park J.J.
      • Kim S.G.
      Expression of transforming growth factor beta1, transforming growth factor type I and II receptors, and TNF-alpha in the mucosa of the small intestine in infants with food protein-induced enterocolitis syndrome.
      presumably because this disease entity is relatively rare in incidence and is encountered during infancy in human subjects but not seen in experimental animals. Here, for the first time, we were able to detect antigen-specific TH2 cell responses in infants with non–IgE-mediated gastrointestinal food allergies by analyzing 89 blood samples collected from all over Japan.
      The antigen-specific lymphocyte stimulation test is a classic method for investigating antigen-specific T-cell proliferation and theoretically should be applicable to the study of gastrointestinal food allergies. However, a couple of previous studies demonstrated that the antigen-specific lymphocyte stimulation test was useful, whereas another study found no such usefulness.
      • Giavi S.
      • Megremis S.
      • Papadopoulos N.G.
      Lymphocyte stimulation test for the diagnosis of non-IgE-mediated cow’s milk allergy: a step closer to a noninvasive diagnostic tool?.
      We hypothesized that this controversy was due to contamination of the antigen preparations with LPS and tested this hypothesis. The limulus amebocyte lysate assay detected high concentrations of LPS in commercially available milk protein preparations, as previously reported (see Table E1 in this article’s Online Repository at www.jacionline.org).
      • Brix S.
      • Bovetto L.
      • Fritsché R.
      • Barkholt V.
      • Frøkiaer H.
      Immunostimulatory potential of beta-lactoglobulin preparations: effects caused by endotoxin contamination.
      In addition, significant lymphoproliferative responses were found in the presence of as little as 10 pg/mL LPS (see Fig E1, A, in this article’s Online Repository at www.jacionline.org), and PBMCs from younger children showed more pronounced lymphoproliferation in response to LPS (see Fig E1, B). Therefore we attempted to remove contaminating LPS from milk protein preparations by passing them through a prepacked endotoxin affinity column. However, a high LPS concentration was detected even after that treatment (see Table E1), and therefore we obtained a special β-lactoglobulin preparation with very low contaminating LPS levels (kindly provided by Bean Stalk Snow, Tokyo, Japan). Further studies were performed by using these milk protein preparations, which contained LPS at a final concentration of less than 5 pg/mL.
      Next, to elucidate what types of antigen-specific immune responses are induced in patients with gastrointestinal food allergies, we cultured PBMCs from patients and control subjects in the presence and absence of LPS-depleted milk component proteins. The study enrolled 65 patients with gastrointestinal food allergies, 12 patients with IgE-mediated cow’s milk allergy (CMA) who showed only nongastrointestinal symptoms on ingestion of milk, and 12 control subjects who showed absolutely no symptoms on ingestion of milk. Table I
      • Powell G.K.
      Food protein-induced enterocolitis of infancy: differential diagnosis and management.
      summarizes the clinical symptoms, clinical diagnosis, and demographic data for the 2 patient groups. None of the patients with gastrointestinal food allergies had detectable levels of IgE against milk proteins in sera. We were unable to recruit infants with IgE-mediated CMA who were age matched with the infants with non–IgE-mediated gastrointestinal food allergies. This study was approved by regional ethics committees, and written informed consent was obtained from the guardians of all patients and control subjects.
      Table IDemographic characteristics of the patients
      IgE-mediated CMAGastrointestinal food allergies
      No.No.
      Age (mo)1238.0 (26.5-60.0)652.0 (1.0-4.0)
      Male/female sex127/56540/25
      Day of onset126532.5 (7.0-115.5)
      Symptoms at onset
       Vomiting120% (0/12)6553.8% (35/65)
       Bloody stool120% (0/12)6547.7% (31/65)
       Diarrhea120% (0/12)6547.7% (31/65)
       Failure to thrive120% (0/12)6538.4% (22/65)
       Lethargy120% (0/12)6538.4% (22/65)
       Fever120% (0/12)6518.5% (12/65)
       Eczema12100% (12/12)657.7% (5/65)
       Wheeze1233.3% (3/12)650% (0/65)
      Laboratory data
       Milk-specific IgE (IU/mL)1256.95 (11.74-90.8)65<0.34 (<0.34)
       Peripheral blood eosinophils (%)Not examined537.7 (3.6-13.5)
      Data are expressed as medians (interquartile ranges). The inclusion criteria were as follows: (1) gastrointestinal symptoms were present more than 2 hours after ingestion of milk and (2) 3 of Powell’s criteria were fulfilled,
      • Powell G.K.
      Food protein-induced enterocolitis of infancy: differential diagnosis and management.
      including (a) switch to therapeutic milk leading to resolution of symptoms, (b) differential diagnosis from other disorders, and (c) verified body weight gain. A definitive diagnosis based on the results of oral food challenge tests that were performed after complete resolution of the initial symptoms was achieved in 19 patients. Patients with gastrointestinal symptoms within 2 hours after ingestion of milk were excluded. On the basis of such symptoms as vomiting, diarrhea, and failure to thrive, the patient group (n = 65) consists of 34 patients with FPIES, 4 patients with food protein–induced enteropathy syndrome (enteropathy), and 27 patients with food protein–induced proctocolitis syndrome (proctocolitis). A definitive diagnosis based on the results of oral food challenge tests was achieved in 13 and 6 patients with FPIES and proctocolitis, respectively. None of the patients underwent endoscopic biopsy.
      The details of the lymphoproliferation test and cytokine production assay are described in the Methods section in this article’s Online Repository at www.jacionline.org. In brief, PBMCs from heparinized peripheral blood were suspended at a cell density of 1 × 106/mL in AIM-V medium (Gibco, Grand Island, NY) without serum. Lymphoproliferation was measured by using tritiated thymidine uptake during a 16-hour period after a 5-day stimulation with 100 μg/mL of each LPS-depleted milk protein preparation (α-lactalbumin, β-lactoglobulin, and α-, β- and κ-caseins). PBMCs were suspended at 1 × 106/mL in RPMI 1640 medium supplemented with 5% autologous plasma to investigate the antigen-specific cytokine production profiles. Culture supernatants were harvested at day 6 after stimulation with 100 μg/mL of each LPS-depleted milk protein preparation, and the cytokine production profiles were investigated by using the Luminex multiplex cytokine analysis kits (Millipore, Bedford, Mass) and ELISA (R&D Systems, Minneapolis, Minn).
      In the first series of experiments, we investigated milk protein–specific lymphoproliferation in the control subjects, patients with IgE-mediated CMA, and patients with gastrointestinal food allergies. The lymphoproliferation level was similar in the patients with IgE-mediated CMA and those with gastrointestinal food allergies. Unlike in previous studies, however, the control subjects showed almost no proliferation (Table II). We presume that this was due to the extensive depletion of LPS contaminating the antigen preparations and the use of serum-free medium.
      Table IIAntigen-specific lymphoproliferation and cytokine production profiles in patients with gastrointestinal food allergies, patients with IgE-mediated allergy, and control subjects
      Control subjectsIgE-mediated CMAGastrointestinal food allergiesP value
      Nonparametric test to compare control subjects and patients with IgE-mediated CMA.
      P value
      Nonparametric test to compare control subjects and patients with gastrointestinal food allergies.
      No.No.No.
      Proliferation (SI)
      The stimulation index (SI) was calculated as milk protein–specific tritiated thymidine uptake (cpm)/vehicle-induced tritiated thymidine uptake (cpm).
      201.290 (0.830-1.738)93.077 (2.484-3.492)652.894 (2.004-7.147)<.01<.001
      Cytokine (pg/mL)
       TNF-α1274.69 (58.44-144.8)1077.78 (58.04-141.4)65241.0 (89.21-729.6)NS<.05
       IL-61279.24 (36.36-193.8)10337.9 (57.43-1021)651151 (157.0-4802)NS<.01
       IL-1β1126.02 (6.880-46.47)1027.49 (6.548-65.04)6448.75 (11.7-136.1)NSNS
       IL-2124.15 (0.0-10.04)1012.31 (7.23-17.58)5816.32 (7.760-39.49)NS<.01
       IL-3120.0 (0.0-0.38)100.40 (0.0-3.61)624.22 (0.0-29.49)NS<.05
       IL-4
      According to the standard curve, the minimal detection limit was 5.88 pg/mL.
      125.365 (2.895-6.358)103.795 (2.033-7.788)655.670 (2.775-12.06)NSNS
       IL-5122.080 (0.0-19.56)1046.59 (4.663-173.5)6563.66 (7.360-310.4)NS<.01
       IL-10129.285 (3.075-15.71)1056.17 (18.74-76.91)6557.92 (12.61-198.8)NS<.05
       IL-131221.61 (0.270-65.04)1082.56 (16.28-555.3)65291.7 (22.10-1417)NS<.01
       IFN-γ113.910 (0.0-67.06)1031.91 (3.635-102.0)6571.86 (5.49-303.4)NSNS
       IL-17120.0 (0.0-2.350)107.635 (1.710-39.63)657.150 (0.0-17.83)NSNS
      PBMCs from each patient were stimulated separately with each of 5 different milk protein preparations, and the data show the highest concentration of each cytokine detected in response to the 5 different stimuli. Data are expressed as medians (interquartile ranges).
      The stimulation index (SI) was calculated as milk protein–specific tritiated thymidine uptake (cpm)/vehicle-induced tritiated thymidine uptake (cpm).
      Nonparametric test to compare control subjects and patients with IgE-mediated CMA.
      Nonparametric test to compare control subjects and patients with gastrointestinal food allergies.
      § According to the standard curve, the minimal detection limit was 5.88 pg/mL.
      In the next experiments we investigated the cytokine production profiles in these subjects. TNF-α concentrations in the culture supernatants of milk protein–stimulated PBMCs from patients with gastrointestinal food allergies were significantly greater than those seen in patients with IgE-mediated CMA or control subjects. However, TNF-α levels in supernatants from patients with IgE-mediated CMA and control subjects were similar (Table II). Significantly higher concentrations of another proinflammatory cytokine, IL-6, were also seen only in the patients with gastrointestinal food allergies.
      The concentrations of 3 TH2 cytokines, IL-3, IL-5, and IL-13, in the supernatants of milk protein–stimulated PBMCs from patients with IgE-mediated CMA tended to be higher than those in the control subjects, but the differences did not reach statistical significance. In contrast, statistically significant and much higher concentrations of these TH2 cytokines were found for the patients with gastrointestinal food allergies. Another TH2 cytokine, IL-4, was undetectable in almost all subjects, and there were no differences among the 3 groups.
      Concentrations of the TH1 cytokine IFN-γ and the TH17 cytokine IL-17 did not show statistically significant differences between any 2 groups.
      The milk component that caused the most prominent tritiated thymidine uptake or the most prominent IL-2 or TNF-α production varied among the patients (see Fig E3 in this article’s Online Repository at www.jacionline.org), suggesting that the lymphoproliferation and cytokine production observed in these assays were indeed antigen specific. In addition, the IL-5 concentration in the culture supernatant of cow’s milk protein–stimulated PBMCs from patients with gastrointestinal food allergies correlated significantly with the peripheral blood eosinophil ratio at disease onset (see Fig E4 in this article’s Online Repository at www.jacionline.org), suggesting that our in vitro assay reflects the in vivo conditions in these patients.
      Collectively, TH2 cytokines, including IL-3, IL-5, and IL-13, but not the TH1 cytokine IFN-γ or the TH17 cytokine IL-17 were significantly produced in vitro by milk protein–stimulated PBMCs from patients with gastrointestinal food allergies. The findings that tritiated thymidine uptake correlated significantly with IL-13 production (data not shown) along with the absence of milk-specific IgE antibody strongly suggest that the IL-13 detected in our assay was not produced by basophils in the PBMC fraction. IL-13 is a well-established mediator of intestinal epithelial cell damage in patients with injuries and inflammatory diseases through activation of the tumor necrosis factor-like weak inducer of apoptosis-fibroblast growth factor-inducible molecule 14 (TWEAK-Fn14) axis.
      • Kawashima R.
      • Kawamura Y.I.
      • Oshio T.
      • Son A.
      • Yamazaki M.
      • Hagiwara T.
      • et al.
      Interleukin-13 damages intestinal mucosa via TWEAK and Fn14 in mice—a pathway associated with ulcerative colitis.
      Thus in addition to the previously known TNF-α, IL-13 might play a crucial role in the pathogenesis of gastrointestinal food allergies.
      In conclusion, antigen-specific T-cell responses in patients with non–IgE-mediated gastrointestinal food allergy are predominantly skewed to TH2. It remains unclear why antigen-specific IgE antibodies were not detected in these patients. Possible explanations are that neonatal B cells scarcely express IL-4/IL-13 receptors
      • Tian C.
      • Kron G.K.
      • Dischert K.M.
      • Higginbotham J.N.
      • Crowe Jr., J.E.
      Low expression of the interleukin (IL)-4 receptor alpha chain and reduced signalling via the IL-4 receptor complex in human neonatal B cells.
      or that production of IgE antibodies had just started but was still undetectable. This question warrants further study.
      We express our sincere gratitude to all the members of the Japanese Research Group for Neonatal Infantile Allergic Disorders. We also thank all the doctors, nurses, and technicians, especially Ms Nao Aida from the Division of Allergy, Gastroenterology, Pathology, Surgery, Interdisciplinary Medicine and Neonatology of the National Center for Child Health and Development, for their hard work and valuable comments. We also thank Professor Mitsuaki Kimura (Department of Allergy and Clinical Immunology, Shizuoka Children’s Hospital) for his valuable suggestions.

      Methods

      Heparinized blood samples were stored at room temperature and transferred to the National Research Institute for Child Health and Development in Tokyo. The following procedures were performed no later than 24 hours after phlebotomy. PBMCs were obtained from peripheral blood by using Ficoll-Hypaque gradient sedimentation (Lymphocyte Separation Medium; ICN Biochemicals, Aurora, Ohio). The viability determined by using trypan blue dye exclusion (Sigma, St Louis, Mo) always exceeded 95%. PBMCs were suspended at a cell density of 1 × 106/mL in AIM-V medium (Gibco) without serum for lymphoproliferation, and in RPMI 1640 medium (GIBCO/Life Technologies, Gaithersburg, Md) in the presence of 5% autologous plasma for cytokine production assays.
      Lymphoproliferation was measured based on tritiated thymidine (Amersham, Tokyo, Japan) uptake during a 16-hour period after 5 days of stimulation with 100 μg/mL of each LPS-depleted milk protein preparation (α-lactalbumin, Sigma; β-lactoglobulin, Bean Stalk Snow; and α-, β-, and κ-caseins, Sigma) at 37°C in a humidified 5% CO2 atmosphere. Incorporated tritiated thymidine was counted with a liquid scintillation counter (TopCount NXT; PerkinElmer Life Sciences, Boston, Mass). The stimulation index was calculated as milk protein–specific tritiated thymidine uptake (cpm)/vehicle-induced tritiated thymidine uptake (cpm).
      Culture supernatants were harvested at day 6, and the cytokine production profiles were investigated by using Luminex multiplex cytokine analysis kits (Millipore) and ELISA (R&D Systems).
      The lymphoproliferation assays and cytokine production assays were performed in duplicates and triplicates, respectively.
      Figure thumbnail fx1
      Fig E1A, LPS at as little as 10 pg/mL can induce lymphoproliferation. PBMCs from young children (n = 60, 0-60 months of age) were stimulated with various concentrations of LPS (Sigma) for 5 days. Lymphoproliferation was measured by using tritiated thymidine uptake. The stimulation index was calculated as milk protein–specific tritiated thymidine uptake (cpm)/vehicle-induced tritiated thymidine uptake (cpm). *P < .05. B, LPS-induced lymphoproliferation was inversely associated with age. PBMCs from young children (n = 21, 0-240 weeks of age) were stimulated with 100 pg/mL LPS (Sigma) for 5 days. Lymphoproliferation was measured by using tritiated thymidine uptake. The stimulation index was calculated as milk protein–specific tritiated thymidine uptake (cpm)/vehicle-induced tritiated thymidine uptake (cpm).
      Figure thumbnail fx2
      Fig E2IL-2 concentrations in culture supernatant of cow’s milk protein–stimulated PBMCs correlated significantly with antigen-specific lymphoproliferation. PBMCs from children with gastrointestinal food allergies were stimulated separately with 100 μg/mL of each of 5 LPS-depleted milk protein preparations in the absence of serum for the antigen-specific lymphoproliferation assay and in the presence of 5% autologous plasma for the IL-2 production assay. The stimulation index was calculated as milk protein–specific tritiated thymidine uptake (cpm)/vehicle-induced tritiated thymidine uptake (cpm), and the highest stimulation index shown among the 5 tested protein preparations was used as that patient’s data in the plot. Even under slightly different culture conditions, antigen-specific lymphoproliferation and antigen-specific IL-2 production were significantly correlated (r = 0.269, P = .025).
      Figure thumbnail fx3
      Fig E3The milk protein component causing the most prominent tritiated thymidine uptake varied among the patients. PBMCs from children with gastrointestinal food allergies were stimulated separately with 100 μg/mL of each of 5 LPS-depleted milk protein preparations in the absence of serum. Lymphoproliferation was measured based on tritiated thymidine uptake. The stimulation index (S.I.) was calculated as milk protein–specific tritiated thymidine uptake (cpm)/vehicle-induced tritiated thymidine uptake (cpm). For each patient, the SI is shown for the PBMCs’ response to each of the 5 milk protein preparations. Each row represents a single patient, and each column represents one of the 5 milk proteins. ALA, α-Lactalbumin; BLG, β-lactoglobulin; blue, SI < 2; yellow, 2.0 < SI < 4.0; red, SI > 4.
      Figure thumbnail fx4
      Fig E4IL-5 concentration in the culture supernatant of cow’s milk protein–stimulated PBMCs correlated significantly with the peripheral blood eosinophil percentage. PBMCs from children with gastrointestinal food allergies were stimulated separately with 100 μg/mL of each of 5 LPS-depleted cow’s milk protein preparations in the presence of 5% autologous plasma for 6 days. Antigen-specific IL-5 production correlated significantly with the peripheral blood eosinophil percentage at disease onset (r = 0.435, P = .007).
      Table E1Concentrations of LPS in commercially available milk protein preparations before and after treatment with a prepacked endotoxin affinity column
      Cow’s milk protein preparationBefore treatment (pg/mg)After treatment (pg/mg)
      α-Lactalbumin (Sigma L-6010)184,20014
      β-Lactoglobulin (Sigma L-3908)206,7001,880
      α-Casein (Sigma C-6780)54023
      β-Casein (Sigma C-6905)50034
      κ-Casein (Sigma C-0406)40041
      LPS-depleted β-lactoglobulin (Bean Stalk Snow)29
      The indicated milk protein preparations were treated with a prepacked endotoxin affinity column (Detoxi-Gel; Pierce Chemical, Rockford, Ill) in accordance with the manufacturer’s instructions. LPS concentrations were measured by using the limulus amebocyte lysate assay.

      References

        • Chung H.L.
        • Hwang J.B.
        • Park J.J.
        • Kim S.G.
        Expression of transforming growth factor beta1, transforming growth factor type I and II receptors, and TNF-alpha in the mucosa of the small intestine in infants with food protein-induced enterocolitis syndrome.
        J Allergy Clin Immunol. 2002; 109: 150-154
        • Giavi S.
        • Megremis S.
        • Papadopoulos N.G.
        Lymphocyte stimulation test for the diagnosis of non-IgE-mediated cow’s milk allergy: a step closer to a noninvasive diagnostic tool?.
        Int Arch Allergy Immunol. 2012; 157: 1-2
        • Brix S.
        • Bovetto L.
        • Fritsché R.
        • Barkholt V.
        • Frøkiaer H.
        Immunostimulatory potential of beta-lactoglobulin preparations: effects caused by endotoxin contamination.
        J Allergy Clin Immunol. 2003; 112: 1216-1222
        • Powell G.K.
        Food protein-induced enterocolitis of infancy: differential diagnosis and management.
        Compr Ther. 1986; 12: 28-37
        • Kawashima R.
        • Kawamura Y.I.
        • Oshio T.
        • Son A.
        • Yamazaki M.
        • Hagiwara T.
        • et al.
        Interleukin-13 damages intestinal mucosa via TWEAK and Fn14 in mice—a pathway associated with ulcerative colitis.
        Gastroenterology. 2011; 141: 2119-2129
        • Tian C.
        • Kron G.K.
        • Dischert K.M.
        • Higginbotham J.N.
        • Crowe Jr., J.E.
        Low expression of the interleukin (IL)-4 receptor alpha chain and reduced signalling via the IL-4 receptor complex in human neonatal B cells.
        Immunology. 2006; 119: 54-62