Association of FceR1-b polymorphisms with asthma andassociated traits in Australian asthmatic families

L. J. PALMER*†, P. J. RYE*, N. A. GIBSON*, M. F. MOFFATT‡, J. GOLDBLATT*,P. R. BURTON*†, W. O. C. M. COOKSON‡ and P. N. LESOUE¨ F*

*Department of Paediatrics, University of Western Australia,†Division of Biostatistics and Genetic Epidemiology,TVW Telethon Institute for Child Health Research, Perth, Australia and‡Nuffield Department of Clinical Medicine,University of Oxford, Oxford, UK

Summary

Background Asthma is a genetically complex disease, and is characterized by elevatedserum immunoglobulin E (IgE) levels, elevated blood eosinophil counts and increasedairway responsiveness. Polymorphisms in theb subunit of the high affinity receptor for IgE(FceR1-b) have been previously associated with these phenotypes and with an increased riskof asthma.Objective To investigate the association of all known bi-allelic polymorphisms in FceR1-bto asthma and quantitative traits associated with asthma in a selected sample of Australianasthmatic children and their nuclear families.Methods Australian Caucasian nuclear families (n¼ 134 subjects) were recruited on thebasis of a child proband with current, severe, symptomatic asthma. The quantitative traitsassessed included serum levels of total IgE and specific IgE to house dust mite and mixedgrass, blood eosinophil counts and the dose–response slope of the forced expiratory volumein 1 s to histamine provocation.Results Neither the Leu181 nor the E237G mutations were detected in this population.Allele B of RsaI intron 2 (RsaI_in2*B) was significantly associated with physician-diagnosed asthma (ever) (P¼ 0.002). Alleles of both theRsaI_in2 and RsaI exon 7(RsaI_ex7) polymorphisms were significantly associated with loge total serum IgE levelsand the combined RAST index.RsaI_ex7was also associated with loge blood eosinophilcounts. These associations were independent of age, sex and familial correlations.Conclusion This study supports a role for the FceR1-b gene or a nearby gene in thepathogenesis of asthma.

Keywords: asthma, bronchial hyperreactivity, eosinophils, FceR1-b, genetics, IgE

Clinical and Experimental Allergy, Vol. 29, pp. 1555–1562. Submitted 13 November 1998;revised 10 May 1999; accepted 2 June 1999.

Introduction

As in other developed nations, asthma is common inAustralia, and is a major public health problem, affectingmore than 20% of the paediatric population and 5–10% ofadults [1–3]. Asthma is closely associated with increasedlevels of total serum immunoglobulin E (IgE) [4], elevatedlevels of IgE specific to common aeroallergens [5] andelevated blood eosinophil counts [6]. Asthma is also typified

by non-specific increased airway responsiveness (AR) toinhaled agents such as histamine or methacholine [7]. Thereis substantial epidemiological and molecular evidence thatboth asthma and these asthma-associated quantitative pheno-types are likely to be determined in part by genetic factors[5,8,9].

While much progress has been made towards definingthe molecular genetics of asthma over the past decade, theintricacy of the numerous genetic and environmental factorsinvolved has complicated genetic dissection of this andother complex diseases [10]. Research into the molecular

Clinical and Experimental Allergy,1999, Volume 29, pages 1555–1562

1555q 1999 Blackwell Science Ltd

Correspondence: L. Palmer, University Department of Paediatrics, PrincessMargaret Hospital for Children, GPO Box D184, Perth 6001, Australia.

genetics of asthma has generally focused on candidate geneswith clearly defined roles in the allergic process [5], such astheb-subunit of the high affinity receptor for IgE (FceR1-b).

Linkage of markers in the 11q13 chromosomal region toserum IgE responses has been reported in several Caucasianand non-Caucasian populations [11–16]. FceR1-b has beenidentified as a potential candidate gene [17] for the observedlinkage. Coding and non-coding polymorphisms within theFceR1-b gene have been related to allergic asthma, atopicdermatitis, serum total IgE levels and other measures ofatopy [18–23].

The aim of this study was to investigate the association ofknown bi-allelic polymorphisms within the FceR1-b geneto asthma and quantitative traits associated with asthma.We investigated the association of four bi-allelic poly-morphisms in the FceR1-b gene with physician-diagnosedasthma, serum total and specific IgE levels, blood eosino-phil counts and AR in a sample of Australian Caucasiannuclear families selected on the basis of an asthmaticproband.

Materials and methods

Study population

Data were collected from a population of families ascer-tained from a respiratory outpatient clinic at the tertiarychildren’s hospital in Perth, Western Australia on thebasis of a Caucasian proband aged 6–15 years, withcurrent, physician-diagnosed, severe, symptomatic asthmarequiring inhaled corticosteroid therapy (n¼ 26 families, 134subjects).

All data collection activities were approved by the EthicsCommittee of the Princess Margaret Hospital for Children.Informed personal and parental consent was obtained for allsubjects participating.

Data collection

Individual and family histories of respiratory symptoms,illnesses and exposures were assessed by modified Amer-ican Thoracic Society questionnaires [24]. Adults wereadministered the questionnaire. Questionnaires relating tochildren were administered to a parent, generally themother. Physician-diagnosed asthma (ever) in subjectswas defined by answers to the questions ‘Have you everhad asthma?’ and ‘Who diagnosed the asthma?’.

Blood was taken by venipuncture from all consentingsubjects (n¼ 131; 98%) for IgE assays and DNA studies.A full-blood picture was performed on each subjectfrom whom blood was taken; this included an automatedwhite cell count to establish a total blood eosinophilcount.

IgE assays

Total serum IgE titres and specific serum IgE titres to wholehouse dust mite (HDM) (Dermatophagoides pteronyssinus;D1) and mixed grass (GX2) were measured using theImmunocap FEIA (Pharmacia AB, Uppsala, Sweden). A‘combined RAST index’ was calculated for each individualas the sum of the RAST scores to HDM and mixed grass.

Airway responsiveness

Spirometry was performed in the standing position using ahand held spirometer (Welch Allyn Model 61000, SkaneatelesFalls, NY, USA) calibrated daily with a 3-L syringe. The bestforced expiratory volume in 1 s (FEV1) was measured accord-ing to the guidelines of the American Thoracic Society [25].

Response to histamine challenge was assessed by theYanet al.rapid method [26]. FEV1 was measured followinginhalation of saline and increasing (doubling) doses ofhistamine delivered from calibrated hand-held nebulizers.At each dose step, one forced expiratory manoeuvre wasperformed and only repeated if the initial manoeuvrewas thought to be technically unsatisfactory. The challengewas continued until either the FEV1 fell by at least 20%from the post-saline value or until the maximal cumulativedose (7.8mmol) of histamine was delivered. AR wasexpressed as the two-point dose–response slope (DRS) ofhistamine response against percentage fall in FEV1 [27].

Molecular analysis

DNA was extracted from whole blood samples by standardphenol-chloroform extraction. The genomic DNA of allindividuals was genotyped using polymerase chain reaction(PCR) techniques for four bi-allelic polymorphisms withinthe FceR1-b gene: a mutation in exon 6 resulting in anleucine–isoleucine substitution within the fourth transmem-brane domain (Leu181) [18], a mutation in exon 7 resultingin a glutamic acid–glycine substitution in the cytoplasmictail (E237G) [21], andRsaIpolymorphisms in intron 2 [28](RsaI_in2) and the untranslated region of exon 7 [22](RsaI_ex7). Genotyping and phenotyping were performedblind with regard to each other.

With the exception of the reactions for the Leu181polymorphism, PCR reactions were performed in sterile96-well thermostable microtitre plates (Costar, Corning,Acton, MA, USA) overlaid with light mineral oil (Sigma,St Louis, MO, USA), and using a Hybaid OmnigeneTMthermocycler utilizing block control.

Detection of the Leu-181 [18] polymorphism in FceR1-bwas undertaken using the amplification refractory mutationsystem (ARMS) PCR [29]. The following oligonucleotideprimers were used: (a)5FU: 50-TGT ATG TGT CAC TTTAAA AGG ACT GGT CAG-30; (b) 5WK: 50-TTG TCA TTT

1556 L. J. Palmeret al.

q 1999 Blackwell Science Ltd,Clinical and Experimental Allergy, 29, 1555–1562

GTT GCT GTT CAA TAG GAA GTT-30; (c) 3M: 50-AATGGT GAG AAA CAG CAT CAT CAT TAC CAA-30; and(d) 3FU: 50-TAA CAT ATC AGT CCT ATT ATC CCAACC CTC-30. PCR reactions were carried out using sterile0.5 mL Eppendorf tubes in 50mL volumes overlaid withlight mineral oil (Sigma). The reaction mix included 300 ngof genomic DNA, 2 U of Taq polymerase (BoehringerMannheim, Mannheim, Germany), 3mL 10xNH4 reactionbuffer (Bioline, Biotech International, British Biotech Phar-maceuticals, Oxford, UK), 200mmol/L of each dNTP (Phar-macia), 0.5mmol/L of each primer except3M (0.1mmol/L)and 1.5 mmol/L MgCl2. Reactions were performed using anMJ Research PTC-100 thermal cycler. A ‘hot-start’ wasused: 40mL of the reaction mixture with DNA but withoutTaq polymerase was heated to 948C for 5 min and held at808C while enzyme was added in the remaining volume ofreaction buffer. Amplification conditions were then 35cycles of 60 s at 948C, 120 s at 558C and 60 s at 728C,followed by a single 10 minute extension cycle at 728C.Three bands potentially resulted [20] from the ARMS PCR: a459-bp control band (5FU-3FU), a 353-bp band in the pre-sence of ‘wild-type’ Il181 (5 WEEK-3FU), and a 163-bp band(3M-5FU) in the presence of Leu181 (designated ‘allele B’).

Detection of the E237G [21] polymorphism was alsoundertaken using ARMS PCR [29]. The following oligonu-cleotide primers were used: (a)B7FA1: 50-TGG CCA GCTAGT CTG GTT TGG TTT TCT GGA-30; (b) B7FA2: 50-GGA GCA TAT TAA GGT GGA CAG AAG CAG CAG-30;(c) B7M1: 50-ATT CAG CTA CTT ACA GTG AGT TGGAAG ACC CAG GCG G-30; (d) B7W2: 50-CAC GTG ATTCTT ATA AAT CAA TGG GAG GAG ACA ATT-30.Genomic DNA samples (50 ng) were amplified in a totalvolume of 50mL containing 50 mmol/L KCl, 10 mmol/LTRIS-HCl (pH 8.3), 200mmol/L of each dNTP (Pharmacia),0.5mmol/L of each primer exceptB7M1 (0.25mmol/L) and1.5 mmol/L MgCl2. A hot-start PCR was done with theaddition of 2 U of Taq polymerase (Bioline) after an initial5 min denaturation at 948C. Reaction conditions were then35 cycles of 1 min at 948C, 2 min at 608C and 2 min at728C, followed by a single 10 min extension cycle at 728C.PCR products were visualized in an ethidium-bromidestained 4% (w/v) agarose gel (3 : 1 NuSieve GTG: LMPagarose; FMC BioProducts, Rockland, ME, USA). Threebands potentially resulted from the ARMS PCR: a 446-bpcontrol band (B7FA1-B7FA2), a 280-bp band in the pre-sence of ‘wild-type’ (B7FA1-B7 W2), and a 238-bp band(B7M1-B7FA2) in the presence of the E237G polymorph-ism (designated ‘allele B’).

RsaI_in2 detection was carried out by PCR with thefollowing oligonucleotide primers: (a) 50-TCT GTC TGTCGA GAA TGT TGC-30, and (b) 50-CTG GTT AGA TCTGAG AAA GAG-30. Genomic DNA samples (150 ng) wereamplified in a total volume of 15mL containing 0.3mmol/L

of each primer, 200mmol/L of each dNTP, 2.0 mmol/LMgCl2, 67 mmol/L TRIS-HCl (pH 8.0), 16 mmol/L(NH4)2SO4, 0.01% TWEEN-20 and 0.5 units ofTaq DNAPolymerase (Bioline UK). Amplification conditions were34 cycles of 948C for 1 min, 598C for 1 min, 728C for 2 min.

RsaI_ex7detection was carried out by PCR with thefollowing oligonucleotide primers: (a) 50-TCA CTG TGTATC ATG CTA AGC-30, and (b) 50-TGA TAC AAT ACTGCA TCG TGG-30. Genomic DNA samples (100 ng) wereamplified in a total volume of 15mL containing 0.5mmol/Lof each primer, 200mmol/L dNTP’s, 1.5 mmol/L MgCl2,67 mmol/L TRIS-HCl (pH 8.0), 16 mmol/L (NH4)2SO4,0.01% TWEEN-20 and 0.75 units DNA Taq Polymerase(Bioline UK). Amplification conditions were 32 cycles of948C for 1 min, 608C for 1 min, 728C for 1 min

After confirmation of successful PCR, 5mL of RsaI_in2or RsaI_ex7PCR product was digested in a final volume of10mL with 0.75 U of RsaIenzyme (New England Biolabs,Beverly, MA, USA) with the recommended buffer for 1 h at378C. Digested products were then separated on 2% agarosegels stained with ethidium bromide. Three banding patternspotentially resulted from the digestion of theRsaI_in2PCRproduct: a 780-bp band (AA genotype), 780 bp, 500 bp and280 bp bands (AB genotype), or 500 bp and 280 bp bands(BB genotype) [28]. An additional, constant 148 bp bandwas present as a result of the primers used. Three bandingpatterns potentially resulted from the digestion of theRsaI_ex7PCR product: a 481-bp band (AA genotype), a481, 295 and 187 bp band (AB genotype), or a 295 and187 bp band (BB genotype) [22].

Statistical analysis

Gender (male¼ 1; female¼ 2) was analysed as a binarycovariate. All other variables were analysed as continuouscovariates. Total serum IgE levels, the blood eosinophilcount and the DRS were all skewed with a long right-handtail, and were loge transformed prior to analysis. Statisticalsignificance was taken at the 5% level.

The primary dichotomous response variable of the asso-ciation analyses was the trait physician diagnosed asthma(ever). Association analysis between this dichotomous traitand each polymorphism was performed using the Transmis-sion Disequilibrium Test (TDT) [30] as implemented inG.A.S. v2.0 (q Alan Young, 1995, Oxford University, UK).

The primary quantitative response variables of the asso-ciation analyses were the following traits: (i) total serum IgEtitre; (ii) combined RAST index to house dust mite andmixed grass; (iii) peripheral blood eosinophil count; and(vii) the two-point DRS to histamine challenge. As age andsex were significantly associated with these phenotypes,they were included as covariates in all models. Age wascentred at its mean and was modelled as a linear, quadratic

Association of FceR1-b polymorphisms with asthma 1557

q 1999 Blackwell Science Ltd,Clinical and Experimental Allergy, 29, 1555–1562

and cubic fixed effect in order to adjust for any non-lineareffects upon outcome.

As a preliminary screening technique, generalized linearmodels (GLMs) were constructed to evaluate associationsbetween the markers investigated and the continuousresponse variables (adjusted for gender and age). Genotypewas modelled as a categorical factor using two dummyvariables in a GLM (AA¼ 0, AB¼ 1, BB¼ 2). Responsevariables showing significant associations with markergenotypes were then subjected to extended analyses usingmaximum likelihood methods to account for familialcorrelations.

Association analysis of the markers investigated withcontinuous outcomes was performed by use of a maximumlikelihood technique that allows for familial correlations ofgenotypes and phenotypic traits [31]. The ASSOC moduleof the software package S.A.G.E. (1994) was used toestimate fixed effects and to partition observed phenotypicvariance into ‘polygenic’ and residual environmental compo-nents using maximum likelihood. Models reported in thispaper invoke the assumption that the distribution of theresponse phenotype is multivariate Normal, with a meanthat depends upon the particular set of explanatory covariatesit includes.

The bi-allelic polymorphisms were coded into threeclasses (AA¼ 0, AB¼ 1,BB¼ 2) and analysed categoricallyas two binary (1,0) dummy variables relative to genotypeAA. ASSOC was also used to estimate genotype specificmean differences and standard deviations relative to abaseline genotype.

The statistical association of covariates entered as fixedeffects and continuous outcomes was formally assessed byremoval of terms from the mean model and calculation ofthe likelihood ratio test statistic. The asymptotic distributionof the test statistic was assumed to be<x2

p when p termswere removed from the model.

Results

The dataset consisted of 26 nuclear families (n¼ 134 sub-jects). The mean number of subjects per family was 5.2(range¼ 4–9). All families contained both parents and atleast two children. The mean age of parents was 41.4 years(SD¼ 6.4 years) and of children 11.9 years (SD¼ 6.5 years).The sex ratio was balanced; 62 male (46.3%) subjects werestudied. Approximately 58% of the population, 40% ofparents and 69% of children, had a physician diagnosis ofasthma. The geometric mean serum total IgE level was112.5 IU/mL, and the mean RAST index was 3.1. Thegeometric mean eosinophil count was 298.6×106 cells, andthe geometric mean DRS to histamine was a 4.4% fall inFEV1 permmol of histamine.

Allele B of RsaI_in2(RsaI_in2*B) comprised 64.8% of

all known alleles; allele B ofRsaI_ex7 (RsaI_ex7*B)comprised 31.5% of all known alleles (Table 1).RsaI_in2and RsaI_ex7 were in tight linkage disequilibrium(x2

4¼ 143.6,P<0.0001); allele A of each polymorphismwas associated with allele B of the other polymorphism.Neither the Leu181 nor the E237G coding variants weredetected in the study population.

Asthma

The TDT analysis indicated thatRsaI_in2*B demonstratedevidence of significant transmission disequilibrium withphysician-diagnosed asthma (ever) (Table 2). Comparisonof maternally and paternally derived alleles did not suggestany significant parent-of-origin effects for eitherRsaI_in2or RsaI_ex7(data not shown).

Quantitative traits

Both RsaI_in2and RsaI_ex7demonstrated evidence of asignificant relationship with loge total serum IgE levels andthe combined RAST index (Table 3).RsaI_in2also showedevidence of a significant relationship with loge blood eosi-nophil levels. ForRsaI_in2, mean adjusted levels of totaland specific serum IgE levels were lowest for theAAgenotype, higher for theAB genotype, and highest for theBB genotype (Table 4). ForRsaI_ex7, mean adjustedlevels of total and specific serum IgE levels and bloodeosinophil counts were lowest for theBB genotype, higherfor the AB genotype, and highest for theAA genotype

1558 L. J. Palmeret al.

q 1999 Blackwell Science Ltd,Clinical and Experimental Allergy, 29, 1555–1562

Table 1.Heterozygosity and proportion of known alleles for typedpolymorphisms

RsaI_in2 RsaI_ex7

Observed heterozygosity 0.504 0.508Alleles known (%) 244/268 (91.0%) 241/268 (89.9%)

Table 2.TDT analysis of FceR1-b polymorphisms with physiciandiagnosed asthma (ever)

Marker Allele Transmitted Not-transmitted P-value*

RsaI_in2 A 21 45 –B 45 21 0.002

RsaI_ex7 A 38 27 0.11B 27 38 –

*P-value derived from exact one-sided binomial distribution.

(Table 5). Neither marker demonstrated evidence of signi-ficant association with the loge DRS to histamine (Table 3).

Extended modelling using the ASSOC module ofS.A.G.E. included age, gender and eitherRsaI_in2 orRsaI_ex7genotypes as covariates. Tables 4 and 5 showthe results of the maximum likelihood modelling of totaland specific serum IgE levels and blood eosinophil counts.Genotype-specific means relative to a baseline genotype,both unadjusted and adjusted for familial correlations, areshown for each outcome.

Both RsaI_in2andRsaI_ex7remained significantly asso-ciated with loge total serum IgE level and the combinedRAST index after adjustment for familial correlations(Tables 4 and 5). Estimation of the familial and non-familial(environmental) components of residual variance [31] indi-cated that the inclusion of theRsaI_in2 (Table 4) andRsaI_ex7 (Table 5) genotypes in the model explainedapproximately 3.8 and 5.9%, respectively, of the totalresidual variance in total serum IgE levels (after adjustmentfor age and gender). Similarly, inclusion of theRsaI_in2(Table 4) andRsaI_ex7(Table 5) genotypes in the model

explained approximately 4.5 and 4.7%, respectively, of thetotal residual variance in RAST index (after adjustment forage and gender).

Based on the results of the linear modelling (Table 3),only the RsaI_ex7polymorphism was investigated furtherfor association with blood eosinophil counts. TheRsaI_ex7polymorphism remained significantly associated with theblood eosinophil count after adjustment for familial correla-tions (Table 5). Estimation of the familial and non-familial(environmental) components of residual variance [31] indi-cated that the inclusion ofRsaI_ex7genotype in the modelexplained approximately 5.4% of the total residual variancein blood eosinophil count (after adjustment for age andgender).

Discussion

This study tested for within-family associations betweenasthma and asthma-associated phenotypes and severalgenetic markers in a selected ‘asthmatic’ population. Asso-ciation studies using polymorphisms within candidate genes

Association of FceR1-b polymorphisms with asthma 1559

q 1999 Blackwell Science Ltd,Clinical and Experimental Allergy, 29, 1555–1562

Table 3. Association of RsaI poly-morphisms and quantitative outcomes RsaI_in2† RsaI_ex7†

Outcome* F statistic P-value F statistic P-value

Loge total IgE level F2,116¼ 6.15 0.003 F2,114¼ 4.78 0.01RAST index F2,116¼ 7.65 0.001 F2,114¼ 5.82 0.004Loge blood eosinophil count F2,112¼ 1.63 0.20 F2,111¼ 5.42 0.005Loge DRS to histamine F2,98¼ 1.20 0.31 F2,96¼ 1.25 0.29

*All models included gender, age, age2, age3. †Genotypes entered as factored variable:AA¼ 0, AB¼ 1, BB¼ 2.

Table 4. Genotype-specific meandifferences forRsaI_in2genotypes andquantitative outcomes

Genotype Genotype-specific Adjusted mean x2

Outcome (N) mean difference (SD)* difference (SD)† (P-value)‡

Loge total serum AA (16) Reference Reference 7.31

IgE AB (50) 1.09 (0.20) 0.74 (0.35) (P¼ 0.03)BB (57) 1.44 (0.19) 1.07 (0.39)

RAST Index AA (16) Reference Reference 7.53AB (50) 2.88 (0.46) 1.92 (0.93) (P¼ 0.02)BB (57) 3.68 (0.43) 2.59 (1.01)

*Genotype-specific mean differences relative toAAgenotype from GLM; adjusted for gender,age, age2, age3.†Genotype-specific mean differences relative toAA genotype from ASSOC model; adjustedfor gender, age, age2, age3, and familial correlations.‡x2 test derived from likelihood ratio statistic (ASSOC model).x2 with two degrees offreedom (marker coded into three classes).

have the potential to identify genetic factors associated withcomplex traits, and are generally more powerful thanlinkage-based approaches [32]. This is particularly true forgenes of modest effect.

The current study was unable to detect the Leu181 orthe E237G coding variants. This suggests that these poly-morphisms may be rare in the families of asthmatic chil-dren, and hence not likely to explain previous linkages ofthe 11q13 chromosomal region to asthma phenotypes.Further, it is possible that inherent methodological problemswith detection might lead to a high false-negative rate forthe Leu181 mutation [33]. Several other published studieshave also failed to detect the Leu181 variant [34].

RsaI_in2*B has been previously been shown to be sig-nificantly associated with both atopic asthma in childhoodand adulthood and ‘atopy’ (raised total or specific serum IgElevels) using a case-control design in a Japanese population(n¼ 600 subjects) [19]. BothRsaI_in2*B andRsaI_ex7*Awere associated with atopic dermatitis and allergic asthmain two samples of British families (n¼ 60 and 88 families)selected on the basis of a child proband with atopicdermatitis [23].RsaI_ex7*A was closely associated withincreased serum IgE levels in a community-based sample(n¼ 234 subjects) of Australian Aboriginal families [22].The results of the current study, which found evidence ofsignificant association betweenRsaI_in2*B or RsaI_ex7*Aand risk of physician-diagnosed asthma, elevated serum IgElevels and elevated blood eosinophil counts, are consistentwith these previous studies.

While the two RsaI markers investigated have beendescribed as being subject to maternal modification in

relation to atopic dermatitis [23], no evidence of a maternaleffect on either marker was found in the current study inrelation to physician-diagnosed asthma.

The high-affinity receptor for IgE on mast cells and baso-phils is a biologically plausible candidate, as it plays animportant role in triggering the allergic reaction [17,35,36].Studies at the cellular level suggest that FceR1-b geneticvariants on mast cells may have a direct regulatory effectupon IgE synthesis through enhanced IL-4 production andmay also make the receptor more sensitive to ligand [37].

The specific function of theb-subunit in triggeringcellular degranulation remains unclear. No mutations inFceR1-b with functional effects that might explain the find-ings of linkage and association of intragenic polymorphismsto asthma-associated phenotypes have been reported. Whileit is possible that the Leu181 and E237G polymorphismshave direct functional effects on asthma-associated pheno-types [20,21], there is no evidence that the twoRsaIpoly-morphisms investigated have any functional significance,particularly as neither is translated.

The results of this study suggest that functional poly-morphism(s) in linkage disequilibrium with alleles of theRsaI_in2andRsaI_ex7markers had a significant regulatoryinfluence on asthma risk, serum IgE levels and bloodeosinophil counts in a sample of selected families. Theestimated proportion of age, gender and familial correla-tion-adjusted total variance in total and specific serum IgElevels and blood eosinophil counts associated with theRsaImarkers was only around 3–6%. This suggests that anyfunctional mutation(s) in the region of FceR1-b had onlymodest effects on IgE- and eosinophil-mediated immune

1560 L. J. Palmeret al.

q 1999 Blackwell Science Ltd,Clinical and Experimental Allergy, 29, 1555–1562

Genotype Genotype-specific Adjusted mean x2

Outcome (N ) mean difference (SD)* difference (SD)† (P-value)‡

Loge total serum AA (62) Reference Reference 8.62IgE AB (46) – 0.06 (0.22) 0.18 (0.26) (P¼ 0.01)

BB (13) – 1.37 (0.41) – 1.02 (0.42)

RAST Index AA (62) Reference Reference 5.90AB (46) – 0.21 (0.48) – 0.14 (0.50) (P¼ 0.06)BB (13) – 3.34 (0.91) – 2.34 (1.09)

Loge blood AA (61) Reference Reference 7.79Eosinophil count AB (45) – 0.06 (0.10) – 0.21 (0.14) (P¼ 0.02)

BB (12) – 0.71 (0.20) – 0.64 (0.22)

*Genotype-specific mean differences relative toAAgenotype from GLM; adjusted for gender,age, age2, age3.†Genotype-specific mean differences relative toAA genotype from ASSOC model; adjustedfor gender, age, age2, age3, and familial correlations.‡x2 test derived from likelihood ratio statistic (ASSOC model).x2 with two degrees offreedom (marker coded into three classes).

Table 5. Genotype-specific mean dif-ferences forRsaI_ex7genotypes andquantitative outcomes

responses, and that any such genetic variants are unlikely tofully account for the substantial heritability of these traitsdescribed in previous studies [8,9,38–40]. This also sug-gests there were multiple genetic determinants of serum IgEand blood eosinophil levels and of asthma risk, and isconsistent with the marked heterogeneity underlyingasthma suggested by numerous other studies [5].

Acknowledgements

We thank the children and parents who participated in thisstudy, and the many colleagues who assisted in the datacollection. We are grateful to Michael Hill of the OxfordAsthma Genetics Group for his assistance with the geno-typing, and to Ingrid Laing for performing the IgE assays.This study was supported by the National Health andMedical Research Council of Australia and the WellcomeTrust (UK). LJP was supported by the British Council (UK)and is a National Health and Medical Research Council ofAustralia Postdoctoral Fellow in Genetic Epidemiology.NAG was supported by a British Medical Associationtravelling fellowship. Some of the results in this paperwere obtained by using the program package S.A.G.E.,which is supported by a USA Public Health Service resourcegrant from the National Center for Research Resources.

References

1 Salome C, Peat J, Britton W, Woolcock A. Bronchial hyper-responsiveness in two populations of Australian school chil-dren. 1. Relation to respiratory symptoms and diagnosedasthma. Clin Allergy 1987; 17:271–81.

2 Woolcock A, Peat J, Salome C. Prevalence of bronchialhyperresponsiveness and asthma in a rural adult population.Thorax 1987; 42:38–44.

3 Forero R, Bauman A, Young L, Larkin P. Asthma prevalenceand management in Australian adolescents: results from threecommunity surveys. J Adolescent Health 1992; 13:707–12.

4 Burrows B, Martinez F, Halonen M, Barbee R, Cline M.Association of asthma with serum IgE levels and skin-testreactivity to allergens. N Engl J Med 1989; 320:271–7.

5 Sandford A, Weir T, Pare P. The genetics of asthma. Am JRespir Crit Care Med 1996; 153:1749–65.

6 Bousquet J, Chanez P, Vignola AM, Lacoste JY, Michel FB.Eosinophil inflammation in asthma. Am J Respir Crit CareMed 1994; 150:S33–8.

7 Ryan G, Latimer K, Dolovich J, Hargreave F. Bronchialresponsiveness to histamine: relationship to diurnal variation ofpeak flow rate, improvement after bronchodilator, and airwayscalibre. Thorax 1982; 37:423–9.

8 Moro-Furlani A, Krieger H. Familial analysis of eosinophiliacaused by helminthic parasites. Genet Epidemiol 1992; 9:185–90.

9 Cookson W, Palmer L. Investigating the Asthma Phenotype.Clin Exp Allergy 1998; 28 (Suppl. 1):88–9.

10 Lander E, Schork N. Genetic dissection of complex traits.Science 1994; 265:2037–48.

11 Young R, Sharp P, Lynch J et al. Confirmation of geneticlinkage between atopic IgE responses and chromosome 11q13.J Med Genet 1992; 29:236–8.

12 Shirakawa T, Hashimoto T, Furuyama J, Morimoto K. Linkagebetween severe atopy and chromosome 11q13 in Japanesefamilies. Clin Genet 1994; 46:228–32.

13 Hizawa N, Yamaguchi E, Furuya K et al. Association betweenhigh serum total IgE levels and D11S97 on chromosome 11q13in Japanese subjects. J Med Genet 1995; 32:363–9.

14 Daniels S, Bhattacharrya S, James A et al. A genome-widesearch for quantitative trait loci underlying asthma. Nature1996; 383:247–50.

15 Boguniewicz M, Hayward A. Atopy, airway responsiveness,and genes. Thorax 1996; 51 (Suppl. 2):S55–S59.

16 Cox N, Beaty T, Rich S et al. Genome screen for asthmasusceptability loci in US populations. Am J Respir Crit CareMed 1998; 157:A855.

17 Sandford AJ, Shirakawa T, Moffatt MF et al. Localisation ofatopy andb subunit of high-affinity IgE receptor (FCeRI) onchromosome 11q. Lancet 1993; 341:332–4.

18 Shirakawa T, Li A, Dubowitz M et al. Association betweenatopy and variants of theb subunit of the high-affinityimmunoglobulin E receptor. Nature Genet 1994; 7:125–30.

19 Shirakawa T, Mao X-Q, Sasaki S, Kawai M, Morimoto K,Hopkin J. Association between FceRIb and atopic disorder in aJapanese population. Lancet 1996; 347:394–5.

20 Hill M, James A, Faux J et al. FceRI-b polymorphism and riskof atopy in a general population sample. Br Med J 1995;311:776–9.

21 Hill M, Cookson W. A new variant of the beta subunit ofthe high-affinity receptor for immunoglobin E (FC-epsilon-RI-beta E237G) — associations with measures of atopy andbronchial hyper-responsiveness. Hum Mol Genet 1996;5:959–62.

22 Palmer L, Pare P, Faux J et al. Fc (Epsilon) R1-Beta Poly-morphism and Total Serum Ige Levels In Endemically Para-sitized Australian Aborigines. Am J Hum Genet 1997; 61:182–8.

23 Cox H, Moffatt M, Faux J et al. Association of atopicdermatitis to the beta subunit of the high affinity immunoglo-bulin E receptor. Br J Dermatol 1998; 138:182–7.

24 Helsing K, Comstock G, Speizer F et al. Comparison of threestandardised questionnaires on respiratory symptoms. Am RevRespir Dis 1979; 120:1221–31.

25 ATS. Standardisation of spirometry, 1994 update (AmericanThoracic Society). Am J Respir Crit Care Med 1995; 152:1107–36.

26 Yan K, Salome C, Woolcock A. Rapid method for measure-ment of bronchial responsiveness. Thorax 1983; 38:760–5.

27 O’Connor G, Sparrow D, Taylor D, Segal M, Weiss S. Analysisof dose–response curves to methacholine: an approach suitableto population studies. Am Rev Respir Dis 1987; 136:1412–7.

28 Shirakawa T, Mao X, Sasaki S et al. Association betweenatopic asthma and a coding variant of FCeRIb in a Japanesepopulation. Hum Mol Genet 1996; 5:1129–30.

Association of FceR1-b polymorphisms with asthma 1561

q 1999 Blackwell Science Ltd,Clinical and Experimental Allergy, 29, 1555–1562

29 Newton C, Graham A, Heptinstall L et al. Analysis of any pointmutation in DNA. The amplification refractory mutationsystem (ARMS). Nucl Acids Res 1989; 17:2503–16.

30 Spielman R, McGinnis R, Ewens W. The transmission/dis-equilibrium test detects cosegregation and linkage [letter]. AmJ Hum Genet 1994; 54:559–60.

31 George V, Elston R. Testing the association between poly-morphic markers and quantitative traits in pedigrees. GenetEpidemiol 1987; 4:193–201.

32 Risch N, Merikangas K. The future of genetic studies ofcomplex human diseases. Science 1996; 273:1516–7.

33 Hill M, Cookson W. Atopy in Australia (reply). Nature Genet1995; 10:260.

34 Thomas N, Holgate S. Genes for asthma on chromosome 11: anupdate [editorial]. Clin Exp Allergy 1998; 28:387–91.

35 Kinet J-P, Blank U, Brini A et al. The high-affinity receptor for

immunoglobulin E. a target for therapy of allergic diseases. IntArch Allergy Appl Immunol 1991; 94:51–5.

36 Humbert M, Grant J, Taborda-Barata L et al. High-affinity IgEreceptor (FceRI) -bearing cells in bronchial biopsies fromatopic and nonatopic asthma. Am J Respir Crit Care Med1996; 153:1931–7.

37 Ravetch J. Atopy and Fc receptors: mutation in the message?Nature Genet 1994; 7:117–8.

38 Gerrard J, Rao D, Morton N. A genetic study of immunoglo-bulin E. Am J Hum Genet 1978; 30:46–58.

39 Hopp R, Bewtra A, Watt G, Nair N, Townley R. Geneticanalysis of allergic disease in twins. J Allergy Clin Immunol1984; 73:265–70.

40 Hanson B, McGue M, Roitman-Johnson B, Segal NL, BouchardTJ, Blumenthal MN. Atopic disease and immunoglobulin E intwins reared apart and together. Am J Hum Genet 1991; 48:873–9.

1562 L. J. Palmeret al.

q 1999 Blackwell Science Ltd,Clinical and Experimental Allergy, 29, 1555–1562