No. 555248; BD Biosciences, San Jose, CA, USA). Statistical testing was performed separately for results before and after challenge. Results are presented as box-plots showing the median, 25th–75th percentile, 10th–90th percentile and outliers. Lupin- and fenugreek-specific IgE antibodies were determined in individual sera at exsanguination by the heterologous PCA-test [25, 26]. In the lupin model, this website we also tested for cross-reactivity by the use of the PCA-test, using legume extracts other than lupin. Briefly, mouse serum

(100 μl) was injected intradermally in rats. Twenty-four hours later, a saline solution containing legume extract (0.1 mg/ml) and Evans Blue (4.5 mg/ml; Sigma-Aldrich, St. Louis, BTK inhibitor MO, USA) was administered i.v. One hour later, the rats were killed and the reactions were read as size in diameter of the blue dots in the skin (illustrations in [25]). All serum samples were diluted 1:4. Prechallenge sera were not available

for PCA because of the relatively large amount of mouse serum needed to perform the test. IgG1 ELISA.  Polystyrene microtiter plates (Maxisorp; VWR International, Radnor, PA, USA) were coated with 2 μg/ml lupin or fenugreek extract and incubated for 2 h at 37 °C and then at 4 °C overnight. Serum samples and antibodies were diluted 1:100 in PBS with 0.05% Tween 20 (PBS-Tw). PBS-Tw was also used as washing buffer between each step. Eight selected serum samples were preincubated with extracts of lupin, fenugreek, peanut, soy or OVA in concentrations from 0 to 10 mg/ml for 1 h to demonstrate the inhibitory potential of the corresponding extracts. All samples were added to the plates and incubated ifenprodil for 2 h at 37 °C. Antibodies were detected by peroxidase-labelled rat monoclonal anti-mouse IgG1 (BD Pharmingen, Franklin Lakes, NJ, USA) for 1 h at 37 °C and peroxidase substrate (ortho-phenylenediamine

chloride; Sigma-Aldrich). Absorbance was determined with an ELISA reader (EL808; BioTek Instruments, Winooski, VT, USA) at 450 nm. Antibody concentrations were given in arbitrary units (AU) per ml. Results are presented as a dose-response curve of the median values and box-plots showing median, 25th–75th percentile, 10th–90th percentile and outliers. Splenocyte preparation.  Spleen cells were prepared by pressing the spleens through a 70-μm cell strainer (BD Labware, Franklin Lakes, NJ, USA). The cell concentrations were determined using a Coulter Counter Z1 (Beckman Coulter Inc., Miami, FL, USA). After incubation in culture medium (RPMI 1640 with L-glutamine, supplemented with 10% foetal bovine serum and 1% streptomycin/penicillin) with or without 50 μg/ml legume extract at 37 °C and 5% CO2 for 5 days, the supernatants were collected and stored at −80 °C awaiting analyses. Trial A (Table 1) was performed to establish the lupin model and was included in the analyses to strengthen the control groups.