420

Food and drugreactions andanaphylax

is

Background: Serologic cross-reactivity among legumes has beendescribed; however, it is rarely clinically significant. In this study3 patients with a history of anaphylaxis to pea are described whosubsequently had symptoms after ingestion of peanut.Objective: We investigated whether the peanut-related symptomswere due to cross-reactivity between pea and peanut proteins.Methods: Peanut-related symptoms were documented accord-ing to case history or double-blind, placebo-controlled foodchallenge results. Skin prick tests were performed, and specificIgE levels were determined for pea and peanut with the CAPsystem FEIA. IgE-binding proteins in pea and peanut wereidentified by using immunoblot analysis. Cross-reactivity wasstudied by means of immunoblot and ELISA inhibition studieswith whole extracts and purified allergens.Results: Peanut-related symptoms consisted of oral symptomsin all patients, with additional urticaria and dyspnea orangioedema in 2 patients. All patients had a positive skin pricktest response and an increased IgE level to pea and peanut.Immunoblotting revealed strong IgE binding to mainly vicilinin pea extract and exclusively to Ara h 1 in crude peanutextract. Immunoblot and ELISA inhibition studies with crudeextracts, as well as purified proteins, showed that IgE bindingto peanut could be inhibited by pea but not or only partiallythe other way around.Conclusion: Clinically relevant cross-reactivity between peaand peanut does occur. Vicilin homologues in pea and peanut(Ara h 1) are the molecular basis for this cross-reactivity. (JAllergy Clin Immunol 2003;111:420-4.)

Key words: Pea, peanut, vicilin, Ara h 1, cross-reactivity

The prevalence of food allergy is estimated to beapproximately 2% in adults and 8% in children.1,2 Themost common offending foods are milk, egg, peanut,soybean, tree nuts, crustaceans, and fish.2 Soybean and

peanut are members of the legume family and are oftenrecognized as allergens.3-6 Adverse reactions to otherlegumes, however, hardly have been described. There area few studies reporting allergic reactions to chickpea,lentil, and lupine.7-10 As observed in earlier studies,patients with peanut allergy have extensive serologiccross-reactivity among members of the legume fami-ly.11,12 However, clinically significant cross-reactivityamong legumes appeared to be rare.13-15 Some reports domention adverse reactions to soybean, lupine, and pea inpatients with peanut allergy.7,8,13,16 In Mediterraneanareas lentils are most commonly implicated in food aller-gy frequently associated with symptoms after ingestionof chickpeas.9 However, only one study, by Moneret-Vautrin et al,8 actually confirmed lupine-related symp-toms caused by cross-reactivity with peanut.

Pea allergy is rare and therefore not extensively inves-tigated, resulting in limited information on pea allergens.Malley et al17,18 demonstrated that only the albumin frac-tion produced positive skin test results in 10 pea-sensitivesubjects, and they later purified a green pea allergen of1.8 kd. Sequence identity among legume proteins hasbeen described previously. Burks et al19 described ahomology of 60% to 65% in amino acid sequencesamong vicilin-like proteins (seed storage proteins) in peaand peanut (Ara h 1), and recently, sequence similaritybetween glycinins in peanut (Ara h 3), soybean, and peaof 62% to 72% was described.20,21

This report describes 3 patients with severe pea-relat-ed symptoms who also had peanut-related symptoms.The aim of this study was to clarify potential cross-reac-tivity between these 2 legumes and to identify the aller-gens involved.

METHODS

Patients

Three patients with a history of severe allergic reactions afteringestion of pea who had peanut-related symptoms were investigat-ed in this study. Specific IgE levels to pea and peanut and total IgE

Food and drug reactions and anaphylaxis

Patients with anaphylaxis to pea canhave peanut allergy caused by cross-reactive IgE to vicilin (Ara h 1)

Marjolein Wensing, MD,a André C. Knulst, MD,a Sander Piersma, PhD,b,c

Francesca O’Kane, MSc,b,c Edward F. Knol, PhD,a and Stef J. Koppelman,PhDa,b,c

Utrecht, Zeist, and Wageningen, The Netherlands

From athe Department of Dermatology/Allergology, University Medical Cen-tre Utrecht, Utrecht; bthe Protein Technology Department, TNO Nutritionand Food Research Institute, Zeist; and cthe Centre for Protein Technolo-gy, TNO-WU, Wageningen.

Supported by TNO Nutrition and Food Research.Received for publication June 7, 2002; revised August 24, 2002, and October

11, 2002; accepted for publication October 16, 2002.Reprint requests: Marjolein Wensing, MD, Department of

Dermatology/Allergology G02.124, University Medical Centre Utrecht,Heidelberglaan 100, 3584 CX Utrecht, The Netherlands.

© 2003 Mosby, Inc. All rights reserved.0091-6749/2003 $30.00 + 0doi:10.1067/mai.2003.61

Abbreviation usedDBPCFC: Double-blind, placebo-controlled food challenge

J ALLERGY CLIN IMMUNOL

VOLUME 111, NUMBER 2

Wensing et al 421

Food

and

dru

gre

actio

ns a

ndan

aphy

laxis

levels were determined by using the CAP system FEIA (Pharmacia& Upjohn, Uppsala, Sweden). Skin prick tests with commercial peaand peanut extracts (ALK-Abelló, Nieuwegein, The Netherlands)were performed and recorded as described by Dreborg and Frew.22

Detailed histories of pea- and peanut-related symptoms wereobtained. For one patient, peanut-related symptoms were less con-vincing, which made a double-blind, placebo-controlled food chal-lenge (DBPCFC) necessary. In short, increasing doses of peanutmeal (30 µg-1 g of peanut protein) hidden in mashed potatoes oroatmeal porridge randomly interspersed with an equal amount ofplacebo doses were administered to the patient. An interval of 30minutes was allowed when no reactions occurred. A challengeresult was regarded as positive when objective reactions occurred orwhen subjective reactions occurred repeatedly (3 times or more)after active doses.

Crude protein extracts

Crude protein extracts from pea and peanut were prepared bymixing 1 g of raw ground pea or peanut with 20 mL of 50 mmol/LTris, pH 8.2, for 2 hours. Mixtures were centrifuged (3000g for 10minutes) to remove insoluble parts. Protein content was measuredby using Bradford analysis, with BSA as a standard.23

Purified proteins

Ara h 1 was purified generally as described previously24,25 withminor modifications and was essentially pure (>95%), as judged byusing a densitometer scan of an SDS-PAGE gel stained withCoomassie Brilliant Blue (Fig 1, lane 4). The N-terminal sequencewas determined and appeared to be identical to the earlier publishedsequence of Ara h 1.25 Purified Ara h 1 was stored at –80°C untiluse. Pea vicilin was purified from raw peas of the Solara variety(Pisum sativum) by using a nondenaturing fractionation procedureaccording to the method of Bora et al.26 Purified vicilin was ana-

lyzed on SDS-PAGE and showed main bands at 50, 33, 30, and 20kd (Fig 1, lane 3), which was similar to results with earlier pub-lished band patterns.27,28 Purified vicilin was freeze-dried andstored at –20°C until use.

SDS-PAGE, immunoblotting, and

immunoblot inhibition

SDS-PAGE was performed by using standard equipment (Bio-Rad, Hercules, Calif) with 15% acrylamide gels (15 × 10 cm). Perlane, 7 µg of crude protein extracts and 1.6 µg of the purified pro-teins was used. Gels were stained with Coomassie Brilliant Blue R-250 dissolved in destaining solution (10% HAc [vol/vol] and 5%methanol [vol/vol] in water). After destaining, gels were scannedwith an ImageMaster DTS (Pharmacia & Upjohn). For Western blot-ting, SDS-PAGE gels were prepared as described above, and the sep-arated proteins were transferred to polyvinyldifluoride sheets(Immobilon-P; Millipore Corp, Bedford, Mass) by using standardtechniques. Membranes were blocked with 3% BSA in wash buffer(50 mmol/L Tris, pH 7.5, containing 0.1% BSA and 0.1% Tween-20)for 1 hour at room temperature. Patients’ sera were diluted 1:120 inwash buffer and applied to the membrane (overnight at room tem-perature). In case of inhibition experiments, increasing concentra-tions of Ara h 1 or vicilin (150-500 and 2-10 µg/mL, respectively)were added to the patients’ serum 1 hour before application to theblot membrane. Bound IgE was detected by using a commercial anti-human IgE conjugated to peroxidase (Diagnostic Product Company,Los Angeles, Calif) and a subsequent staining reaction for peroxi-dase activity by using the ECL technique. Between each step, blotmembranes were washed thoroughly 5 times with wash buffer. Non-specific binding of the anti-IgE antibody conjugate was measured ina similar blotting procedure, omitting the incubation step withpatient sera, and appeared to be negligible.

FIG 1. SDS-PAGE and CBB stain: immunoblotting and blot inhibition with serum of patient 3. Lanes 1 to 4,CBB stain of crude pea and peanut extract and vicilin and Ara h 1; lanes 5 to 8, immunoblotting of crudeextracts and the purified allergens; lanes 9 to 12, IgE binding to vicilin and Ara h 1 inhibited by using 2 con-centrations of vicilin; lanes 13 to 16, IgE binding to vicilin and Ara h 1 inhibited by using 2 concentrationsof Ara h 1.

422 Wensing et al J ALLERGY CLIN IMMUNOL

FEBRUARY 2003

Food and drugreactions andanaphylax

is

ELISA techniques

ELISA techniques were based on an earlier report about the Arah 1–IgE interaction.29 Plates (96-well) were coated with 10 µg/mLAra h 1 or pea vicilin by using PBS as a coating buffer. Subse-quently, wells were blocked with BSA (1% in PBS containing 0.1%Tween-20, block buffer) to diminish the nonspecific binding.Patients’ sera were diluted in block buffer, and serum dilutions wereincubated on the allergen-coated ELISA plates for 2 hours at 37°C.In inhibition experiments patient serum was preincubated withinhibitor in increasing concentrations (0-100 µg/mL) for 1 hour atroom temperature. These mixtures were transferred to the allergen-coated ELISA plates and incubated for 2 hours at 37°C. Bound IgEwas detected by using an anti-human IgE antibody conjugated tohorseradish peroxidase (Diagnostic Product Company) and isexpressed as A490. Between each step, plates were washed 5 timeswith PBS containing 0.1% Tween-20. Nonspecific reactivity wasstudied by using serum from a nonallergic individual and appearedto be negligible. Uncoated plates incubated with patient serum didnot bind IgE either, indicating that nonspecific binding of IgE toplastic was negligible too.

RESULTS

Patients

Patient characteristics are summarized in Table I. Pea-related symptoms consisted of symptoms of itching andtingling of the oral cavity or lips and dyspnea in allpatients. Two patients (nos. 1 and 3) experienced addi-tional symptoms, such as generalized urticaria and car-diovascular symptoms, resulting in faintness. Thisrequired attendance at an emergency department morethan once. Ingestion of peanut also induced oral symp-toms in all patients: in patient 1 this was accompanied bygeneralized urticaria and dyspnea and in patient 3 bysevere angioedema, both requiring medication, such asβ-agonists or antihistamines. Patient 2 had oral symp-toms exclusively that were confirmed with a DBPCFC.She reported itching and tingling of the oral cavity froma dose of 100 mg of peanut protein (about two thirds ofa peanut). This also occurred after the 2 following dosesof 300 and 1000 mg but not after placebo doses. Allpatients had severe allergic reactions to pea before thedevelopment of peanut-induced reactions. Two patientsreported additional symptoms to other members of thelegume family. All patients frequently had allergic reac-tions caused by accidental ingestion of legumes hidden inprepackaged foods, such as bread and snacks.

Identification of allergens

Immunoblotting was performed with crude proteinextracts from pea and peanut to investigate the pea andpeanut allergens recognized by IgE of these patients.Fig 1 (lane 5) shows IgE binding with serum of patient3 to crude pea extract. IgE-binding protein bandsappeared to have molecular weights of 50 kd, 33 kd, 30kd, between 20 and 30 kd, and 20 kd or less. Such a pat-tern is indicative for vicilin.27,28 Indeed, immunoblot-ting with purified vicilin, as shown in Fig 1, lane 7,demonstrated similar IgE-binding bands. Yet 2 uniden-tified IgE-binding bands (35 and <14 kd) in crude peaextract could not be explained by IgE to vicilin. Fig 1also shows IgE binding to crude peanut extract. Inter-estingly, only one band stained clearly for IgE. Themolecular weight of this band is approximately 65 kd,which is indicative of Ara h 1.19,24,25 This was furtherconfirmed by immunoblotting with purified Ara h 1(Fig 1, lane 8). Weak IgE binding was also detected toa protein band of approximately 50 kd. Because thisband is also present in purified Ara h 1, this bindingmight be due to a proteolytic fragment of Ara h 1. Takentogether, IgE binding to vicilins in both pea and peanut(Ara h 1) could be detected in this patient. Similarrecognition profiles were demonstrated by using sera ofpatients 1 and 2 (not shown), although the Ara h 1 bandfor patient 2 was less intense, possibly because of a lowpeanut-specific IgE titer (Table I).

Assessment of cross-reactivity of individual

pea and peanut allergens

Cross-reactivity between the homologous vicilin stor-age proteins of the legume family was studied by usingIgE-binding inhibition studies with purified pea vicilinand Ara h 1. Fig 1 shows the IgE reactivity for patient 3under different conditions. Preincubations with increas-ing concentrations of vicilin showed complete inhibitionof IgE binding to both vicilin and Ara h 1. Furthermore,the serum was preincubated with 150 or 500 µg/mL Arah 1. IgE binding to vicilin was still detectable withapproximately the same intensity as in the control situa-tion, even with the highest concentration of Ara h 1. Thebinding to Ara h 1 itself was only partially inhibited by150 µg/mL Ara h 1 but was inhibited in a more pro-nounced manner by 500 µg/mL Ara h 1.

TABLE I. Patient characteristics

CAP Age of Peanut- CAP Age of Total Other legumes

Patient Pea-related SPT pea onset related SPT peanut onset IgE that caused

no. Sex Age (y) symptoms pea (kU/L) (y) symptoms peanut (kU/L) (y) (kU/L) symptoms

1 F 40 Os, urt, dys, ND >100 3 Os, urt, dys 3+ 39 5 352 Mp, kb, hbgi, rc, shock

2 F 32 Os, dys, rc 4+ 21.1 29 Os (confirmed 2+ 1.56 30 216 Le, fb, gb, bs,by DBPCFC) bb, mp

3 M 28 Os, urt, dys, ND 70 3 Os, ae ND 10 27 308 —shock

Os, Oral symptoms; urt, urticaria; dys, dyspnea; gi, gastrointestinal symptoms; rc, rhinoconjunctivitis; ND, not done; mp, marrow pea; kb, kidney bean; hb,haricot bean; le, lentils; fb, French beans; gb, green beans; bs, bean sprouts; bb, broad beans; ae, angioedema.

J ALLERGY CLIN IMMUNOL

VOLUME 111, NUMBER 2

Wensing et al 423

Food

and

dru

gre

actio

ns a

ndan

aphy

laxis

Inhibition of IgE binding was studied in an inhibitionELISA to further substantiate this cross-reactivity and toobtain more quantitative results. Sera were preincubatedwith various concentrations of purified pea vicilin or Arah 1. Fig 2 demonstrates data of patient 3. Fig 2, A, showsthat IgE binding to vicilin can be inhibited by vicilinitself, whereas the vicilin-IgE interaction could not beinhibited by Ara h 1. Fig 2, B, shows that IgE binding toAra h 1 could be inhibited by high concentrations of Arah 1 itself, but the inhibition with vicilin appeared to be farmore potent. Table II summarizes inhibitory concentra-tion of 50% values for IgE inhibition with pea vicilin andAra h 1 in all 3 patients. In all patients IgE binding tovicilin could be inhibited by vicilin itself but not by Arah 1, and the IgE binding to Ara h 1 could be inhibited byAra h 1 itself but far more potently by vicilin.

DISCUSSION

Three patients with histories of severe anaphylaxis afteringestion of pea were investigated regarding their peanut-related symptoms. In one patient peanut-related symptomswere confirmed by means of DBPCFC. The other 2 patientswere not challenged with peanut because they had convinc-ing histories of peanut anaphylaxis (Table I). Looking at thespecific IgE levels, skin prick test results, and the course ofdevelopment of food-related symptoms in these patients, itis assumed that their peanut-related symptoms were due tocross-reactive IgE initially raised against pea allergens.

Remarkably, patients 1 and 2 were raised in an agricul-tural environment, where different legumes were cultivat-ed. Possibly their sensitization to pea is the result of earlyand frequent contact with legumes (skin contact, inhala-tion, and ingestion) during different stages of ripening.Although early sensitization might have occurred, clinicalsymptoms did not always develop at a young age. A sim-ilar situation might be the case in Mediterranean coun-tries, where lentils and chickpeas are widely consumedand have been described to be more frequently involvedin food allergy compared with in Northern Europe.9

Immunoblotting and IgE inhibition assays were used toidentify the involved allergens and to study cross-reactiv-ity in the described patients. Vicilin-like proteins in peaand peanut (Ara h 1) were identified as underlying aller-gens causing cross-reactivity. Although band patterns ofpea vicilin and Ara h 1 differ substantially on SDS-PAGEand immunobloting, the proteins are homologous to alarge extent.19 Pea vicilin consists of a number ofpolypeptides that all originate from a single precursorprotein of 50 kd after posttranslational proteolysis.28 Theposttranslational proteolysis of Ara h 1 is less extensive,resulting in a single band on SDS-PAGE.25

The observation that IgE binding to Ara h 1 could beinhibited by pea vicilin but IgE binding to pea vicilincould not be inhibited by Ara h 1 suggests that pea acts asthe sensitizing agent in these patients, as is also reflectedin the course of symptom development in these patients.

Two patients also reported symptoms to otherlegumes. These findings suggest that sensitization to pea

vicilin induces broad cross-reactivity among members ofthe legume family, which is, at least partly, clinically sig-nificant. Burks et al19 previously described a homolo-gous sequence of 60% to 65% among vicilin-like pro-teins in peanut (Ara h 1) and pea and speculated that theextensive serologic cross-reactivity among members ofthe legume family is partly due to IgE directed againstvicilin-like proteins. However, patients described in ear-lier studies usually had adverse reactions to only onefood of the legume family.13-15 These studies mainlyconcerned persons with peanut allergy. Other studies(from Mediterranean countries) concerned patients aller-gic to lupine, lentil, or chickpea, and they did mentionadverse reactions to other legumes.7-10 Therefore perhapsthe sensitizing legume is of importance in determiningclinically significant cross-reactivity among legumes.

FIG 2. IgE ELISA inhibition studies with purified pea vicilin andAra h 1. Plates were coated with purified pea vicilin (A) or Ara h 1(B) and, after blocking, incubated with diluted serum from patient3 that was preincubated with various concentrations of pea vicilin(squares) or Ara h 1 (circles).

TABLE II. Concentrations of inhibitors necessary toinhibit initial IgE binding by 50%

Ara h 1–IgE interaction Vicilin-IgE interaction

IC50 for IC50 for IC50 for IC50 for

Patient Ara h 1 vicilin Ara h 1 vicilin

no. (µg/mL) (µg/mL) (µg/mL) (µg/mL)

1 18 <0.01 >100 0.1602 >100 0.1 >100 0.0333 76 0.027 >100 0.050

IC50, Inhibitory concentration of 50%.

A

B

424 Wensing et al J ALLERGY CLIN IMMUNOL

FEBRUARY 2003

Food and drugreactions andanaphylax

is

The 3 patients described in this study were sensitized toAra h 1. Ara h 1, identified as a major peanut allergen in1991, is a 64.5-kd glycoprotein that belongs to the vicilinfamily of seed storage proteins.24 Recombinant Ara h 1 wasproduced in 1995.19 There is some diversity in observedprevalences of Ara h 1 recognition in patients with peanutallergy, varying between 35% and 100%.24,25,30,31 It isunclear whether these differences can be explained by theuse of purified versus recombinant Ara h 1 or by differ-ences in patient characteristics. If peanut sensitization canbe explained by cross-reactive IgE in some patients, geo-graphic differences in allergen recognition patterns can beexpected as a result of differences in consumption patternsor other environmental factors.

The potency of Ara h 1 to inhibit the IgE–Ara h 1 inter-action on Western blotting is poor in the case of all 3patients because 150 to 500 µg/mL Ara h 1 is required toobtain partial inhibition. A similar inhibition experiment(not shown) with serum from a patient with peanut allergywho was not sensitized to pea demonstrated that theIgE–Ara h 1 interaction was partially inhibited with 1µg/mL Ara h 1 and completely inhibited with 3 µg/mL Arah 1. These concentrations are in the same range as neededfor vicilin to obtain partial (2 µg/mL) or complete (10µg/mL) inhibition. The fact that this concentration isapproximately 100 times lower than that observed for thepresently investigated patients is remarkable and can beexplained by the fact that these patients are sensitized topeanut through cross-reactive IgE. Nevertheless, thepatients have clear peanut-related symptoms, illustratingbiologic activity of this IgE–Ara h 1 interaction. Perhapsthe less severe reactions described by the patients afteringestion of peanut compared with pea-related symptomsis explained by this different IgE–Ara h 1 interaction. Twopatients, however, mentioned worsening of their peanut-related symptoms, and therefore an increase in the severityof peanut-related symptoms over time is not unthinkable.

In summary, 3 patients with a history of anaphylaxis topea who had peanut-related symptoms were described.The peanut-related symptoms could be explained bymeans of cross-reactive IgE initially directed against pea.The molecular basis for this cross-reactivity pattern wasformed by vicilin homologues in pea and peanut (Ara h 1).

We thank W. J. Koers for his help in recruiting patients with peaallergy and R. Vlooswijk and R. van Biert for their assistance inWestern blot and ELISA experiments.

REFERENCES

1. Niestijl-Jansen J, Kardinaal A, Huijbers G, Vlieg-Boerstra B, Martens B,Ockhuizen T. Prevalence of food allergy and intolerance in the adultDutch population. J Allergy Clin Immunol 1994;93:446-56.

2. Bruijnzeel-Koomen C, Ortolani C, Aas K, Bindslev-Jensen C, Björkstén B,Moneret-Vautrin D, et al. Adverse reactions to food. Allergy 1995;50:623-35.

3. Sampson H. Peanut anaphylaxis. J Allergy Clin Immunol 1990;86:1-3.4. Loza C, Brosteff J. Peanut allergy. Clin Exp Allergy 1995;25:493-502.5. Hourihane JO’B, Kilburn SA, Nordlee JA, Hefle SL, Taylor SL, Warner

JO. An evaluation of the sensitivity of subjects with peanut allergy to verylow doses of peanut protein: a randomized, double-blind, placebo-con-trolled food challenge study. J Allergy Clin Immunol 1997;100:596-600.

6. Burks AW, Brooks JR, Sampson HA. Allergenicity of major componentproteins of soybean determined by enzyme-linked immunosorbent assay

(ELISA) and immunoblotting in children with atopic dermatitis and pos-itive soy challenges. J Allergy Clin Immunol 1988;81:1135-42.

7. Hefle SL, Lemanske RF, Bush RK. Adverse reaction to lupine-fortifiedpasta. J Allergy Clin Immunol 1994;94:167-72.

8. Moneret-Vautrin DA, Guerin L, Kanny G, Flabbee J, Fremont S, MorissetM. Cross-allergenicity of peanut and lupine: The risk of lupine allergy inpatients allergic to peanuts. J Allergy Clin Immunol 1999;104:883-8.

9. Pascual CY, Fernandez-Crespo J, Sanchez-Pastor S, Padial A, Diaz-PenaJM, Martin-Munoz F, et al. Allergy to lentils in Mediterranean pediatricpatients. J Allergy Clin Immunol 1999;103:154-8.

10. Martinez San Ireneo M, Ibanez Sandin MD, Fernandez-Caldas E, MaranonLizana F, Rosales Fletes MJ, Laso Borrego MT. Specific IgE levels to Cicerarietinum (chick pea) in tolerant and non-tolerant children: evaluation ofboiled and raw extracts. Int Arch Allergy Immunol 2000;121:137-43.

11. Barnett D, Bonham B, Howden MEH. Allergenic cross-reactions amonglegume foods—an in vitro study. J Allergy Clin Immunol 1987;79:433-8.

12. Eigenmann PA, Burks AW, Bannon GA, Sampson HA. Identification ofunique peanut and soy allergens in sera adsorbed with cross-reacting anti-bodies. J Allergy Clin Immunol 1996;98:969-78.

13. Bock SA, Atkins FM. The natural history of peanut allergy. J Allergy ClinImmunol 1989;83:900-4.

14. Bernhisel-Broadbent J, Sampson HA. Cross-allergenicity in the legumebotanical family in children with food hypersensitivity. J Allergy ClinImmunol 1989;83:435-40.

15. Bernhisel-Broadbent J, Taylor S, Sampson HA. Cross-allergenicity in thelegume botanical family in children with food hypersensitivity. II. Labo-ratory correlates. J Allergy Clin Immunol 1989;84:701-9.

16. Foucard T, Malmheden Yman I. A study on severe food reactions in Swe-den—is soy protein an underestimated cause of food anaphylaxis? Aller-gy 1999;54:261-5.

17. Malley A, Baecher L, Mackler B, Perlman F. The isolation of allergensfrom the green pea. J Allergy Clin Immunol 1975;56:282-90.

18. Malley A, Baecher L, Mackler B, Perlman F. Further characterization ofa low-molecular weight allergen fragment isolated from the green pea.Clin Exp Immunol 1976;25:159-64.

19. Burks AW, Cockrell G, Stanley JS, Helm RM, Bannon GA. Recombinantpeanut allergen Ara h 1 expression and IgE binding in patients withpeanut hypersensitivity. J Clin Invest 1995;96:1715-21.

20. Rabjohn P, Helm EM, Stanley JS, West CM, Sampson HA, Burks AW, etal. Molecular cloning and epitope analysis of the peanut allergens Ara h3. J Clin Invest 1999;103:535-42.

21. Beardslee TA, Zeece MG, Sarath G, Markwell JP. Soybean glycinin G1acidic chain shares IgE epitopes with peanut allergen Ara h 3. Int ArchAllergy Immunol 2000;123:299-307.

22. Dreborg S, Frew A, ed. Allergen standardisation and skin tests. EAACIposition paper. Allergy 1993;48(suppl 14):49-82.

23. Bradford MM. A rapid and sensitive method for the quantification ofmicrogram quantities of protein utilizing the principle of protein-dyebinding. Anal Biochem 1976;72:248-54.

24. Burks AW, Williams LW, Helm RM, Connaughton C, Cockrell G,O’Brien TJ. Identification of a major peanut allergen, Ara h I, in patientswith atopic dermatitis and positive peanut challenges. J Allergy ClinImmunol 1991;88:172-9.

25. de Jong EC, van Zijverden M, Spanhaak S, Koppelman SJ, Pellegrom H,Penninks AH. Identification and partial characterization of multiplemajor allergens in peanut proteins. Clin Exp Allergy 1998;28:743-51.

26. Bora PS, Brekke CJ and Powers JR. Heat induced gelation of pea (Pisumsativum) mixed globulins, vicilin and legumin. J Food Sci 1994;59:594-6.

27. Gatehous JA, Croy RRD, Morton H, Tyler M, Boulter D. Characterisa-tion and subunit structures of the vicilin storage proteins of pea (Pisumsativum L.). Eur J Biochem 1981;118:627-33.

28. Gatehouse JA, Lycett GW, Croy RRD, Boulter D. The post-translationalproteolysis of the subunits of vicilin from pea (Pisum sativum L.).Biochem J 1982;207:629-32.

29. Koppelman SJ, Bruijnzeel-Koomen CAFM, Hessing M, De Jongh HHJ.Heat induced conformational changes of Ara h 1, a major peanut allergen,do not affect its allergenic properties. J Biol Chem 1999;274:4770-7.

30. Kleber-Janke T, Crameri R, Appenzeller U, Schlaak M, Becker WM.Selective cloning of peanut allergens, including profilin and 2S albumins,by phage display technology. Int Arch Allergy Immunol 1999;119:265-74.

31. Clarke MC, Kilburn SA, Hourihane JO, Dean KR, Warner JO, Dean TP. Sero-logical characteristics of peanut allergy. Clin Exp Allergy 1998;28:1251-7.