Patients who complained of respiratory allergy symptoms were recruited during one year from the Allergy Clinic of Ajou University Hospital in Suwon, Korea. Among the 1,738 patients, ELISA was performed on the sera from 56 patients showing ≥2+ positive reactions to chestnuts on skin prick test for detecting serum specific IgE to homegrown chestnuts. Among them, the sera from 9 patients showing high specific IgE levels on ELISA were used for the ELISA inhibitions testing and for IgE-immunoblotting.

Chestnut food allergen extracts were prepared according to the method described previously (8). Briefly, fresh chestnuts (the most commonly ingested one in this country, 1:5 wt/vol) were homogenized in phosphate-buffered saline (PBS, pH 7.5). The homogenate was kept at 4℃ overnight followed by centrifugation at 3,500 rpm for 40 min. The supernatant was dialyzed (the cut-off molecular weight was 6,000 Da; Spectrum Medical Industries, Houston, TX, U.S.A.) against 4 L of PBS at 4℃ for 48 hr. The supernatant was then passed through a Nalgen syringe filter (Rochester, New York, NY, U.S.A.) and lyophilized until ELISA, ELISA inhibition, immunoblotting analysis, and digestibility tests were performed. For skin prick test, the 1:5 wt/vol extract was mixed with an equal amount of glycerin (Sigma Co., St. Louis, MO, U.S.A.).

Routine skin prick tests were performed for 30 food allergens using 26-gage needles on the backs of the patients. The results were read 15 min later. The size of the wheal produced by each allergen and histamine was expressed as the mean diameter of maximum length and vertical length at the midportion of the maximal length. Skin reactivity was expressed as the ratio of the wheal size of the allergen to the histamine (A/H). If the A/H ratio was from 0.1 to 1, but the erythema was >21 mm, it was read as 2+. If the A/H ratio was 1 to 2, it was read as 3+. If the A/H ratio was from 2 to 3, it was read as 4+. If the A/H ratio was greater than 3, it was read as 5+. A positive was defined as equal to or greater than 2+ on skin prick test (9).

ELISA was performed to determine the presence of specific IgE to chestnut as previously described (10). Microtiter plates (Costar, New York, NY, U.S.A.) were coated with chestnut extract at 100 µL/well (diluted in carbonated buffer, pH 9.6, 10 µg/well) and the plates were left at 4℃ for 24 hr. Each well was washed three times with 0.05% PBS-Tween (PBST), and was blocked by incubation with 3% BSA-PBST at 200 µL/well for 1 hr at room temperature (RT). The wells were then incubated for 2 hr at RT with either 50 µL of the patient's sera or the control sera (all the sera was diluted 1 to 5). After washing three times with PBST, biotin-labeled goat antihuman IgE antibody (Vector Co, Burlingame, CA, U.S.A.) was added to the each well and incubated for 1 hr. After washing, the wells were incubated with 100 µL of 1:1,000 vol/vol streptavidin-peroxidase (Sigma Co., St. Louis, MO, U.S.A.) for 30 min before ano- ther washing step. The colorimetric reaction was developed with B (3,3',5,5'-tetramethylbenzidine) substrate solution for 10 min at RT. The reaction was stopped by the addition of 100 µL of 2 N sulfuric acid and the absorbance was read at 450 nm by an automated microplate reader (Benchmark, Bio-Rad, Hercules, CA, U.S.A.). All assays were performed in duplicate. The amounts of specific antibodies in the samples were calculated from control curves that were created with the optical densities from serial dilutions of the positive control samples that showed higher antibody titer by using a spline fit program (Microplate Manager, Bio-Rad), and they were expressed as arbitrary units (A.U.). The chestnut-specific IgE was considered positive if the absorbance value was higher than the cutoff value, which was derived from the mean+3×S.D. of the absorbance values from the 15 control subjects.

Electrophoresis in a 10% polyacrylamide separation gel with 4% stacking gel was performed and the proteins were stained with colloidal blue (Novex, San Diego, CA, U.S.A.) or the proteins were transferred onto a PVDF membrane (Millipore, Bedford, MA, U.S.A.), and this was then incubated with the pooled serum of patients with high serum specific IgE that had been diluted 1:5 vol/vol with PBST. Bound specific IgE was detected by biotin-labeled goat antihuman IgE antibody (Sigma). This was followed by a treatment with 0.66% 4-nitroblue tetrazolium and 0.33% 5-bromo-4-chloro-3-indolyl-phosphate (Sigma), which was used as a substrate.

The fresh chestnut extracts were prepared by heating the crude extracts at 100℃ for 5 min. The digestibility of the crude extract per se and the digestibility of the preheated extract that was incubated in simulated gastric fluid (SGF) and simulated intestinal fluid (SIF) were examined as previously described (4). Briefly, for the SGF digestibility test, a protein sample (680 µg of naive crude protein or the preheated extract) was dissolved in 200 µL of prewarmed 100 mM/L HCl (pH 1.2), along with 30 mM/L NaCl containing a 0.32 wt/vol percentage of pepsin A (Sigma Chemical Co). Digestion was carried out at 37℃ with continuous shaking, and aliquots of this digest solution (20 µL) were withdrawn at 0, 0.5 and 60 min. These aliquots were quickly mixed with 26 µL of a sample buffer (containing 2.5% 2-mercaptoethanol and 1% SDS) together with 6.0 µL of Na₂CO₃ solution (200 mM/L). The mixture was then boiled for 5 min and stored at -20℃ until further analyses. For the evaluation of SIF and SGF, a sample (680 µg of the naive crude protein or the preheated extract) was dissolved in 260 µL of prewarmed intestinal control solution (0.05 M KH₂SO₄, pH 6.8) containing a 1.0 wt/vol percentage of pancreatin (Pancreatin USP, Sigma). This was then incubated at 37℃ with continuous shaking. Aliquots of this digest solution (26 µL) were withdrawn at 1, 90 and 240 min and quickly boiled for 5 min after mixing it with 26 µL of a sample buffer (containing 2.5% 2-mercaptoethanol and 1% SDS), and then 12% SDS-PAGE and IgE immunoblot analysis were repeated as described above.

The results of the skin prick test revealed that 56 (3.2%) of 1,738 patients showed more than 2+ A/H response to the chestnut extracts, as is shown in Table 1. Among them, 15 (0.9%) showed more than 3+ A/H response. The prevalence of serum specific IgE, as determined by ELISA, to the chestnut extracts were 34.1% in the 2+ reactors, 58.3% in the 3+ reactors and 0% in the 4+ reactors. The absorbance value of specific IgE to chestnut, according to skin reactivity, is shown in Fig. 1.

The ELISA inhibition test for wild chestnut using the pooled sera from subjects with high serum specific IgE level showed significant dose-dependent inhibitions by the chestnut and chestnut pollen extracts, and both extracts had similar potencies, while negligible inhibition was noted for D. pteronyssinus, as is shown in Fig. 2.

To observe the protein components of the fresh chestnut extracts, they were analyzed by using 10% SDS-PAGE and IgE-immunoblotting with the patients' sera. As shown in Fig. 3, 21 IgE-binding components were detected within the chestnut extracts, while no bindings were noted with the control sera. The 90, 83, 76, 69, 56, 49, 43, 24 and 13 components were noted in more than 50% of the sensitized patients (Fig. 4).

Two dimensional SDS-PAGE (Fig. 5A) and immunoblotting (Fig. 5B) of the fresh chestnut extracts using the pooled sera of two patients with high levels of specific IgE to chestnut extract showed the 24 kDa IgE binding component and its iso-electric point was 6.3, while a 29 kDa component was not detectable. Table 2 shows the result of amino acid sequencing for this major allergen having a 24 kDa of molecular weight, which has homology with legume protein from oak trees (11).

Immunoblotting using the pooled sera showed that most of the IgE-binding components were substantially digested in 10 to 30 min by SGF in the preheated and heated conditions. After the treatment with SIF, although most of the proteins within the chestnut extracts were not changed, IgE binding components were suppressed in the preheated and heated conditions (Fig. 6, 7).