FULL PAPER Genome screen in the French EGEA study: detection of linked regions shared or not shared by allergic rhinitis and asthma

(5) 6, 95 & 5 Nature Publishing Group All rights reserved 466-4879/5 $. www.nature.com/gene FULL PAPER Genome screen in the French EGEA study: detection of linked regions shared or not shared by allergic rhinitis and asthma M-H Dizier, E Bouzigon, M Guilloud-Bataille,CBétard, J Bousquet 4, D Charpin 5, F Gormand 6, J Hochez 7, J Just 8, A Lemainque, N Le Moual 9, R Matran, F Neukirch, M-P Oryszczyn 9, E Paty, I Pin, D Vervloet 4, F Kauffmann 9, M Lathrop, F Demenais and I Annesi-Maesano 9 INSERM U55, Hôpital Paul Brousse, Bâtiment Leriche, Villejuif Cedex, France; INSERM EMI -6, Evry, France; Centre National de Génotypage, Evry, France; 4 Clinique des maladies respiratoires, INSERM U454, Hôpital Arnaud de Villeneuve, Montpellier, France; 5 Service de pneumologie-allergologie, Hôpital Nord, Marseille, France; 6 Service de Pneumologie, centre hospitalier Lyon-Sud, Pierre Benite; 7 INSERM U46, Paris, France; 8 Centre de diagnostic et traitement de l asthme, Hôpital Trousseau, Paris, France; 9 INSERM U47, Villejuif, France; Laboratoire d exploration fonctionnelle, Hôpital Calmette, Lille, France; INSERM U48, Paris, France; Service de pneumologie infantile, Hôpital Necker, Paris, France; Département de Médecine Aigüe Spécialisée, CHU Michallon, Grenoble, France; 4 Service de pneumo-allergologie, Hôpital Ste Marguerite, UPRES EA 87, Marseille, France In the sample of 95 French EGEA families with at least one asthmatic subject, a genome screen was conducted to identify potential linkage regions specific either to allergic rhinitis (AR) or to asthma as well as those shared by the two diseases. Two binary rhinitis phenotypes based on () diagnosis (AR bin ) and () symptoms (AR bin ) and a categorical ordered trait (AR cat ) were considered. Asthma phenotype was based on answers to a standardized questionnaire plus the presence of bronchial hyper-responsiveness. Linkage analyses were conducted using the maximum likelihood binomial (MLB) method. These analyses provided potential evidence for linkage to three regions in the whole sample: p for the phenotype defined by AR bin plus asthma (P ¼.6), q for AR bin (P ¼.6) and p4 p4 for AR cat (P ¼.). Two other regions were detected in the subset of 85 families with at most one asthmatic sib: 9p and 9q q4 for AR bin (P ¼. and.7, respectively). No region showed evidence for linkage to asthma without being also linked to AR. While p may contain a genetic determinant common to asthma and AR, q, p4 p4, 9p and 9q q4 are more likely to harbor genetic factors specific to AR. (5) 6, 95. doi:.8/sj.gene.6646 Published online 7 January 5 Keywords: allergic rhinitis; asthma; linkage analysis; genome screen Introduction Allergic rhinitis (AR) is clinically defined as a symptomatic disorder of the nose induced by an IgE-mediated inflammation of the nasal membranes. Although AR is not a severe disease, it alters the social life of patients and affects school performance and work productivity. Asthma is a common respiratory disease characterized by variable airflow obstruction, inflammation of the airways and bronchial hyperresponsiveness (BHR). The presence of a genetic component for both AR and asthma has been well established by family studies and twin studies. Asthma and AR are common co-morbidities: most asthmatics have rhinitis and many patients with rhinitis suffer from asthma. Familial association has Correspondence: Dr M-H Dizier, INSERM U55, Hôpital Paul Brousse, Bâtiment Leriche, BP, 9487 Villejuif Cedex, France. E-mail: dizier@vjf.inserm.fr Received 9 July 4; revised December 4; accepted December 4; published online 7 January 5 been shown between AR and asthma. 4 AR may result from genetic determinants common to asthma, but there may be also genetic alterations specific to each of these phenotypes, as indicated by twin studies. 5,6 As for asthma, AR is strongly associated with atopy, with most subjects having rhinitis being atopic, while only a part of atopic subjects suffer from rhinitis. Genetic determinants shared by asthma and AR are likely to result in part from the atopic component of these two diseases. Many genome-wide searches and association studies have been conducted for asthma and atopy, measured by response to skin prick tests or IgE levels and for other related traits including eosinophil counts and bronchial responsiveness and have led to detect numerous linked regions. 7 The regions most often reported to be linked to asthma and/or associated phenotypes are 5q, 6p, p, q and q. These genome screens have led to identification of four genes by positional cloning: ADAM associated with asthma and BHR located on chromosome, 8 PHF associated with IgE levels and severe asthma on chromosome, 9 DPP associated

96 with asthma on chromosome and GPRA associated with high serum IgE or asthma on chromosome 7p. In contrast, only two genome-wide searches, have been recently conducted for AR that showed potential linkage to the following chromosomal regions: 4q4 q7 in a Danish sample of families and p6, 4q and 9q4 in a Japanese sample of 48 families. No linkage region was common to these two studies and only the p6 region had been previously reported for asthma. 4 In a first genome-wide search in 7 French families of the EGEA study (Epidemiological study on the Genetics and Environment of Asthma) with at least two asthmatic siblings, 5 linkage analyses were conducted for asthma, IgE, skin test response, eosinophil counts and bronchial responsiveness using a panel of 54 microsatellites. A second genome scan was recently carried out for asthma/atopy-associated phenotypes in an extended sample of 95 EGEA families (having two sibs with DNA available and at least one asthmatic subject) and using a larger panel of 96 markers. 6 Using the same family sample and panel of markers of this second scan, the aim of the present study was to investigate the genetic factors specific either to AR or to asthma as well as those shared by AR and asthma. Three phenotypes were considered for AR: two binary phenotypes based on self-reporting () diagnosis (AR bin ) and () symptoms (AR bin ) according to a standardized questionnaire and a categorical ordered trait (AR cat ) based on a score computed from various items related to AR. 7 The asthma phenotype was based on answers to a questionnaire plus the presence of BHR. Linkage analyses were conducted using the maximum likelihood binomial (MLB) method, 8 developed for linkage analysis of binary and ordered categorical traits in sibships. Linkage analyses were first applied to the whole sample of 95 families (including 66, 86 and sibships with at least two affected sibs for AR bin, AR bin and asthma, respectively, and 95 sibships informative for AR cat ). These analyses were also conducted for the three AR phenotypes, in a subsample of 85 families including at most one asthmatic sib. This study represents the third published genome scan for AR applied to a larger sample than the two previous scans and the first genome-wide screen for both AR and asthma. These two diseases were analyzed separately and together (AR associated with asthma) to characterize the shared and unshared linked chromosomal regions by these two diseases. Results Distribution of AR and asthma In the 76 genotyped siblings of the 95 EGEA families selected through one asthmatic subject, the respective frequencies of AR bin and AR bin phenotypes were equal to 9 and 4% (44% in males for either phenotype and 4 and 4% in females for AR bin and AR bin, respectively), while 5% of the siblings were asthmatic (Table ). There was a strong relationship of AR bin and AR bin with asthma (Po.) with a large part of sibs with AR having also asthma (8% for AR bin and 74% for AR bin ) compared to a proportion of 6 and 9% of asthmatic among the sibs without AR bin and AR bin, respectively (Table ). Table Distribution of the two allergic rhinitis phenotypes (AR bin and AR bin ) and asthma in 76 genotyped siblings from 95 EGEA families Frequency (%) All Male Female AR bin 9 44 4 AR bin 4 44 4 Asthma 5 58 46 Frequency (%) of asthma among subjects AR bin + AR bin OR (95% IC) AR bin + AR bin OR (95% IC) 8 6 7.9 (5.5.4) 74 9 4. (. 6.) In the 76 genotyped siblings, the Score For Allergic Rhinitis (SFAR) ranged between and 6 with a mean of.7.4 (s.d.), (.7.5 in males and.87.47 in females). Departure from normality remained very strong, even after log-transformation, leading to analyze this score as a categorical trait. Distribution of the ordered categorical trait AR cat, given by the proportions of sibs in the seven ordered categories (defined in the Material and methods section), was.7,.,.,.4,.,.,.. Mann Whitney rank test showed significant association between AR cat and asthma (Po.). Distributions of the 95 EGEA families according to the number of genotyped sibs available for linkage analysis of the six phenotypes investigated (asthma, AR bin,ar bin,ar cat,ar bin and AR bin associated with asthma) are shown in Table. To provide a summary indicator of the amount of information available for analysis of each phenotype, the numbers of independent sib-pairs deduced from the sibship size distribution (a sibship of size n corresponds to (n ) independent sib-pairs) are also given in Table. In the subsample of 85 families with at most one asthmatic sib, these distributions are given for rhinitis phenotypes only. Linkage analysis Linkage analyses were performed for the following affection statuses: /asthma, /AR bin, /AR bin, 4/AR cat, 5/AR bin associated with asthma and 6/AR bin associated with asthma in the whole sample of 95 families and for the AR phenotypes only in the subset of 85 families. Peaks of lod MLB scores leading to a P-value p.5 are shown in Tables (whole sample) and 4 (subset). For the chromosomes including the regions meeting our selection criteria, the curves of lod MLB scores are presented in Figures (whole sample) and (subset). In the whole sample of 95 families (Table ), three regions lead to a score lod MLBX.96 (Pp ): p for AR bin plus asthma (P ¼.6 at the DS89 marker), this region being also potentially linked but less significantly to either AR bin (P ¼.5) or asthma (P ¼.5) at DS797, which is centimorgan () apart from DS89; q for AR bin alone (P ¼.6) and p4 p4 for AR cat alone (P ¼.). Two other regions were also detected with a less stringent criterion (Pp.5): p4 for AR bin

Table Distribution of families according to the number of genotyped sibs used for linkage analyses to asthma and AR phenotypes in the 95 families (85 families) 97 Nb of sibs per family 4 5 Total nb of families Nb of independent sibpairs Asthma 94 5 7 AR bin 55 (4) 9 () () 66 (4) 79 (4) AR bin 77 (7) 8 () 4 () 86 (4) 5 (4) AR bin +asthma 6 6 4 48 AR bin +asthma 4 4 47 5 AR cat 89 (9) 8 (47) 7 (7) 6 () 95 (85) 44 () Table Results of linkage analyses of asthma and rhinitis phenotypes using the MLB method in the whole sample of 95 families Marker Position a Asthma AR bin AR bin AR cat AR bin +asthma AR bin +asthma DS797 p (75 ).5 b (.5).8 (.).5 (.5).4 (.) DS89 p (86 ). (.7). (.). (.).8 (.6) DS64 q (86 ).8 (.6). (.) DS77 p4 p4 (6 ).9 (.) DS94 p4 (6 ).6 (.). (.) DS65 q ( ).9 (.) a Cytogenetic location of linked regions and in parentheses the distance from pter retrieved from http://www.marshmed.org/genetics. b score and in parentheses the P-value. (P ¼.) and q for AR bin (P ¼.). For asthma, only one region, p, was detected with a P-value p.5, this region being also linked to AR. Indeed, no region was detected for asthma alone, that is, without being also linked to one of the rhinitis phenotypes. In the subsample of 85 families without asthmatic sibpairs (Table 4), two other regions were detected with a P-value p. for AR bin alone: 9p (P ¼.) and 9q q4 (P ¼.7). The q and p4 p4 regions, detected in the whole sample, also showed indication of linkage (P ¼.5) in this subsample. As seen from Figures and, AR cat showed different linkage patterns from those obtained with AR bin and AR bin for all chromosomes, while peaks of lod MLB scores for AR bin and AR bin were observed at the same position (although at different levels) for chromosomes p, q and p. Interestingly, when adding asthma to AR in the definition of the affection status, only lod MLB scores in the p region are increased while there is a decrease in lod MLB scores in all other regions. Discussion Although asthma and AR are strongly correlated and have some common biological components including the atopic response, they also differ in many respects such as target organs that present substantial differences (there are no smooth muscles in the upper airways) and variations in their geographical distribution and epidemiological associations. 9 This leads to assume the existence of both common genetic components and specific factors for AR and asthma. The present genome screen showed potential linkage for AR and/or asthma (Pp.) to five regions: three regions being detected in the whole sample (p more significantly for AR bin plus asthma than for AR bin or asthma alone, q for AR bin and p4 p4 for AR cat while two other regions were revealed in the subset of 85 families with at most Table 4 Results of linkage analyses of rhinitis phenotypes using the MLB method in the 85 families without asthmatic sibpairs Marker Position a AR bin AR bin AR cat DS5 q (76 ).5 (.5) DS8 p4 p (9 ).5 (.5) D9S57 9p (5 ).9 b (.) D9S68 9q q4 (8 ). (.7) a Cytogenetic location of linked regions and in parentheses the distance from pter retrieved from http://www.marshmed.org/ genetics. b Z MLB score and in parentheses the P-value. one asthmatic sib for AR bin only (9q q4 and 9p). Indication of linkage (Pp.5) was observed in two additional regions in the whole sample: p4 for AR bin and q for AR bin. No region was found linked to asthma without being also linked to AR. Heterogeneity of results for the different rhinitis definitions, according to whether they are based on diagnosis or presence of symptoms, could be due to the difference of sample sizes used for each phenotype and/or to the type of information brought by each one. Moreover, power of linkage analysis for each type of phenotypes may depend on the underlying genetic model. Note that we used here quite stringent thresholds (Pp. or.5) to select linked regions, because of the increase of type one error due to multiple testing. None of our results reached the genome-wide significant level (Pp 5 ) proposed by Lander and Kruglyack, although results of the two regions (p and q) reached the criteria of suggestive linkage (Pp7 4 ). However, the criteria proposed by Lander and Kruglyack apply to very dense mapping and if the mapping is less dense (as it is the case here), these criteria become conservative. We then conducted whole-genome

98 Chromosome 5 5 5 arbin arbin arcat information content.8.6.4. Chromosome 5 5 arbin arbin arcat information content.8.6.4. 5 5 5 asthma arbin+asthma arbin+asthma information content Chromosome.8.6.4. 5 5 5 arbin arbin arcat information content.8.6.4. 5 5 5 asthma arbin+asthma arbin+asthma information content.8.6.4. 5 5 asthma arbin+asthma arbin+asthma information content 5 5 75 5 5 arbin arbin arcat information content Chromosome 5 5 75 5 5 asthma arbin+asthma arbin+asthma information content.8.6.4..8.6.4..8.6.4. Chromosome 5 45 6 75 9 5 arbin arbin arcat information content.8.6.4. Figure.8.4. 5 45 6 75 9 5 asthma arbin+asthma arbin+asthma information content curves and information content () for five chromosomes with lod MLB scores X.5 in the 95 families..6 simulations to get genome-wide P-values, which were p.5 in two regions, p for AR bin plus asthma and q for AR bin supporting the evidence of linkage to these regions. In contrast, genome-wide P-values were greater than.5 for the three (or five) other selected regions showing a pointwise P-values p. (or.5). Thus, only indication of linkage can be suggested for these latter regions. Note that most published genome scans did not report genome-wide P-values and the published pointwise P-values were, in most instances, of the same order of magnitude as those obtained in the present study. However, accumulating indication for linkage across studies may be more important than unreplicated high linkage peaks to identify regions of interest. The present study is the third published genome scan of AR after one screen conducted in a Danish sample and another one in a Japanese one. The 9q q4 region detected for potential linkage to AR in our sample was also reported in the Japanese scan. The linkage peak in the q region observed here for AR is about 4 apart from the linked region indicated in the Danish scan. The other regions, p6 and 4q, detected in the Japanese genome scan and 4q4 7 detected in the Danish screen were not replicated here. Lack of replication between the results of the French, Danish and the

Chromosome 5 5 5 arbin arbin arcat information content Chromosome 5 5 arbin arbin arcat information content.8.6.4..8.6.4. 99 Figure Chromosome 9 5 5 75 5 5 arbin arbin arcat information content and information content () curves for three chromosomes with lod MLB scores X.5 in the 85 families..8.6.4. Japanese studies may be explained by either statistical issues (spurious detection of linkage, due to multiple tests, and/or lack of power) or by the heterogeneity between samples in terms of the phenotype definition (based only on questionnaire in the EGEA and Japanese studies and based on both questionnaire and clinical examination in the Danish study) and/or by the mode of ascertainment of the families. Moreover, the definition of rhinitis in Japanese and Danish studies included the presence of atopy defined either by RAST or SPT test. We then repeated all linkage analyses of AR by including a positive SPT for at least one allergen in the affection status definition. However, this led to unchanged results (results not shown) or to less significant results than for AR or AR plus asthma phenotypes not taking into account atopy. The ascertainment was performed through subjects with clinical atopy in the Danish study, through two children with orchard grass-sensitive rhinitis in the Japanese study and through asthmatic subjects in the EGEA sample. Thus, these differences in the modes of ascertainment of these samples might explain the different of results. Note that the mode of ascertainment of the French data had the advantage to allow examination of the shared and unshared linked chromosomal regions by asthma and AR. In contrast, most of the regions detected in the present study were previously reported for asthma and/or related phenotypes by other genome scans which did not consider the AR phenotype and were applied to family samples mostly ascertained through asthmatics as in our study. Indeed, the p region has been already reported for linkage to asthma in the Caucasian population of the CSGA study, to atopy and IgE 4 in the Hispanic group of the same study and to IgE in a Dutch population. 5 The q region was previously detected for IgE in the two German 6 and Dutch 5 genome scans. It was also found to be linked to IgE in our sample but less significantly than to rhinitis and neither to skin prick test responses nor to other atopy-related phenotypes. 6 The linkage peak observed here for AR is around 5 apart from the CTLA-4 gene (cytotoxic T lymphocyte antigen 4), a gene that has been shown to be associated with IgE levels, 7 atopy and asthma. 8 Thus, detection of linkage in our study could correspond to the influence of this gene. Further fine mapping of these data will permit to confirm this point. The p4 p4 region has been reported for atopic dermatitis in a Swedish population, 9 asthma and BHR in the Hutterite population 4, and mite sensitivity in the CSGA families. The 9q q4 region has been reported for asthma in the Hutterites, 4

for IgE and atopy in the German scan 5 and the region p4 for SPT and IgE in the Hutterite population, 4, in the first EGEA scan 5 and in the German scan. 5 Finally, the last region detected here for rhinitis, q qter, has been already reported for asthma in the CSGA sample and for Dermatophagoides pteronyssinus-specific IgE responsiveness in Caucasian CSGA subjects. This region is adjacent to the q region detected for eosinophil counts in our sample. 5,6 Results of the present study support evidence of linkage to AR and/or asthma for two regions p and q and suggest indication of linkage to p4 p4, 9p, 9q q4 p4 and q regions. These results do not indicate any linkage specific to asthma. They suggest that p may contain a genetic factor common to asthma and AR while the other regions are more likely to harbor genetic factors specific to AR. Since there is no indication of linkage of the p region to atopy-related phenotypes in our sample (total IgE, SPT, specific IgE), 6 this result is likely not due to the common allergic component of asthma and rhinitis. This is supported by a recent population-based study, which showed that the strong association between asthma and rhinitis was not fully explained by atopy. A hypothesis already proposed at a physio-pathologic level is that asthma and rhinitis are different expressions of the same disease, thus the factor located on p may be involved in the different manifestations of the disease. Another proposed hypothesis is that asthma associated with AR corresponds to a severe form of the same disease; detection of linkage to p would correspond to a genetic factor involved in this severe form. Since there was no or only a weak indication of linkage of this region to AR in the 85 families without asthmatic sib-pairs (P ¼.) and to asthma after excluding the sib-pairs with rhinitis (P ¼.6), the second hypothesis appears more likely than the first one. Further family studies of asthma and AR may be of interest to validate our findings. Material and methods Family data The main EGEA sample included 48 nuclear families selected through one asthmatic proband. An additional sample of 4 nuclear families was selected directly through two asthmatic sibs in order to increase the number of informative families for linkage analyses of asthma. Inclusion criteria met by probands were described elsewhere. 4 As compared to our first genome scan conducted in a subsample of 7 EGEA families having at least two asthmatic sibs with DNA available, 5 the present scan was extended to 7 EGEA families with at least two sibs with DNA available. After excluding families with insufficient DNA available or showing non-mendelian transmission, the analyzed sample consisted of 95 families including 7 individuals. Ethical committee approved the protocol and subjects signed informed consent forms. Phenotypes analyzed rmation on respiratory and allergic symptoms, medical history and environmental factors was collected with a standardized questionnaire. 5 Different clinical tests were performed, which were already described, 5 and bronchial responsiveness was included, which was measured using a protocol described in detail elsewhere. 4 All analyzed phenotypes were based on questionnaire and complementary test information. Asthma phenotype was defined as used for the selection of the 7 families of the first genome scan, 5 with the following criteria: a positive response to at least one of the two questions (() Have you ever had attacks of breathlessness at rest with wheezing? () Have you ever had an asthma attack?) associated with either the presence of BHR (defined as a fall in baseline FEV, the forced expiratory volume at one second, X% at p4 mg/ml methacholine), an increase of baseline FEV X% after bronchodilator use, hospitalization for asthma in life or asthma therapy. 4 Three definitions of AR were considered, all based on questionnaire information. The first was a binary trait based on diagnosis: a positive answer to at least one of the two questions (() Have you ever had AR? or () Have you ever had hay fever?) (AR bin ). The second binary trait was defined by the presence of at least one of the following symptoms: sneezing or runny nose to hay/flower or animals or dust (AR bin ). The third one was derived from the SFAR, which is based on these various items: nasal symptoms in the past year including sneezing, runny nose and blocked nose when the subject did not have a cold or the flu in the past year. nasal symptoms accompanied by itchy-watery eyes (rhinoconjonctivitis). months of the year in which nasal symptoms occurred. triggers of nasal symptoms including pollens, house dust mites, moulds and epithelia. allergic status assessed by a positive answer to the question have you ever been desensitized for allergy in your life? previous medical diagnosis of hay fever. The SFAR is a quantitative score previously validated using doctor s diagnosis of rhinitis and skin prick test positivity as gold standard. 7 The SFAR was shown to be highly correlated with objective markers of immediate hypersensitivity such as specific IgE and skin prick test positivity. 4 For the purpose of the present genetic study, familial resemblance of rhinitis was not introduced in the computation of the score. Genotypes A total of 7 subjects were genotyped with a panel of 96 microsatellites (78 autosomal markers). The Linkage Marker Set MD (Applied Biosystems, Foster City, CA, USA) formed the core marker set for the genome-wide screen. These microsatellite markers, labelled with fluorescent dyes (FAMt, HEXt, NEDt), are distributed at an average marker density of (roughly every million bases in the genome) and have an average heterozygosity of 75%. CNG (Centre National de Génotypage) has developed a protocol allowing the coamplification of up to six of these markers in a single reaction to be robust using dual 84-well GeneAmp s PCR 97 cyclers (Applied Biosytems) and an automated procedure for PCR and purification setup. Automatic genotyping was performed based on a series of software processes implemented in Genetic Profiler software (version.).

Before statistical analysis, rigorous genotype quality assurance was performed to ensure accurate binding of alleles. Consistency of the data with Mendelian inheritance and lack of double recombination between loci was evaluated using Pedcheck 6 and other purpose-written software. Linkage analysis methods Linkage analyses were performed by a model-free approach, the MLB method which can be applied to the whole sib-ship of affected subjects. 8,7,8 In brief, the principle of this method is based on the binomial distribution of the number of affected sibs receiving a given parental allele from heterozygous parents. The probability a for an affected sib to receive from his/her parent the marker allele transmitted with the disease allele is equal to.5 under the null hypothesis of no linkage and a is less than.5 under the hypothesis of linkage. Test for linkage is performed using a likelihood ratio test statistic, L ¼ Ln[L(a)/L(a ¼.5)] with the statistic L being distributed asymptotically as a mixture distribution of.5 w df and.5 w df. The test can also be expressed either as a one-sided standard normal deviate denoted as Z MLB ¼ L / or as a lod score criterion, lod MLB ¼ L/(Ln()). This method, which has been extended to categorical traits, 9 is basically the same as the one proposed for binary traits, with the introduction of a latent binary variable capturing the linkage information between the observed categorical trait and the marker. As a large proportion (.4) of individuals had an SFAR score equal to, the score derived from SFAR could not be analyzed as a quantitative trait in linkage analysis, but as an ordered categorical traits (AR cat )by the MLB method. Seven categories were considered and individuals were classified according to their score being equal to,, ( or ), (4 or 5), (6 or 7), 8 or (9 or greater) respectively. The choice of these classes was motivated by sample sizes within each class. All linkage analyses were conducted using the program MLBGH, 4 which incorporates the MLB method into the multipoint approach of Genehunter. 4 These analyses were first conducted in the whole sample of 95 families (see family data section) and then, for the three AR phenotypes, in a subsample of 85 families in which all families with at least two asthmatic sibs were excluded. This latter sample was considered to distinguish between the regions linked specifically to AR from those linked to asthma, provided, in the whole sample, most AR sibpairs were also asthmatic sibpairs. EGEA cooperative group Respiratory epidemiology: INSERM U47, Villejuif: I Annesi- Maesano, F Kauffmann (co-ordinator), MP Oryszczyn; INSERM U48, Paris: M Korobaeff, F Neukirch. Genetics: INSERM EMI 6, Evry: F Demenais; INSERM U55, Villejuif: MH Dizier; INSERM U9, Paris: J Feingold; CNG, Evry: M Lathrop. Clinical Centers: Grenoble: I Pin, C Pison; Lyon: D Ecochard (deceased), F Gormand, Y Pacheco; Marseille: D Charpin, D Vervloet; Montpellier: J Bousquet; Paris Cochin: A Lockhart, R Matran (now in Lille); Paris Necker: E Paty, P Scheinmann; Paris-Trousseau: A Grimfeld. Data management: INSERM ex-u55, Paris: J Hochez; INSERM U47: N Le Moual. References Bousquet J, Van Cauwenberge P, Khaltaev N. Rhinitis and its impact on asthma. J Allergy Clin Immunol ; 8 (5 Suppl): S47 S4. Dold S, Wjst M, von Mutius E, Reitmeir P, Stiepel E. Genetic risk for asthma, rhinitis, and atopic dermatitis. Arch Dis Child 99; 67: 8. Rasanen M, Laitinen T, Kaprio J, Koskenvuo M, Laitinen LA. Hay fever a Finnish nationwide study of adolescent twins and their parents. Allergy 998; 5: 885 89. 4 Annesi-Maesano I, Cotichini R, Stazi MA. Early gene environment interaction into asthma and rhinitis comorbidity. Chest ; : 755. 5 Hopper JL, Hannah MC, Macaskill GT, Mathews JD. Twin concordance for a binary trait: III. A bivariate analysis of hay fever and asthma. Genet Epidemiol 99; 7: 77 89. 6 Duffy DL, Martin NG, Battistutta D, Hopper JL, Mathews JD. Genetics of asthma and hay fever in Australian twins. Am Rev Respir Dis 99; 4 (Part ): 5 58. 7 Wills-Karp M, Ewart SL. Time to draw breath: asthmasuceptibility genes are identified. Nat Genet 4; 5: 76 87. 8 Van Eerdewegh P, Little RD, Dupuis J et al. Association of the ADAM gene with asthma and bronchial hyperresponsiveness. Nature ; 48: 46 4. 9 Zhang Y, Leaves NI, Anderson GG et al. Positional cloning of a quantitative trait locus on chromosome q4 that influences immunoglobulin E levels and asthma. Nat Genet ; 4: 8 86. Allen M, Heinzmann A, Noguchi E et al. Positional cloning of a novel gene influencing asthma from chromosome q4. Nat Genet ; 5: 58 6. Laitinen T, Polvi A, Rydman P et al. Characterization of a common susceptibility locus for asthma-related traits. Science 4; 4: 4. Haagerup A, Bjerke T, Schoitz PO, Binderup HG, Dahl R, Kruse TA. Rhinitis a total genome-scan for susceptibility genes suggests a locus on chromosome 4q4 q7. Eur J Hum Genet ; 9: 945 95. Yokouchi Y, Scibasaki M, Noguchi E et al. A genome-wide linkage analysis of orchard grass-sensitive childhood seasonal rhinitis in Japanese families. Genes Immun ; : 9. 4 Ober C, Tsalenko A, Parry R, Cox NJ. A second-generation genomewide screen for asthma-susceptibility alleles in a founder population. Am J Hum Genet ; 67: 54 6. 5 Dizier MH, Besse-Schmittler C, Guilloud-Bataille M et al. Genome screen for asthma and related phenotypes in the French EGEA study. Am J Respir Crit Care Med ; 6: 8 88. 6 Bouzigon E, Dizier MH, Krähenbühl C et al. Clustering patterns of LOD scores for asthma-related phenotypes revealed by a genome-wide screen in 95 French EGEA families. Hum Mol Genet 4; :. 7 Annesi-Maesano I, Didier A, Klossek M, Chanal I, Moreau D, Bousquet J. The score for rhinitis (SFAR): a simple and valid assessment method in population studies. Allergy ; 57: 7 4. 8 Abel L, Alcais A, Mallet A. Comparison of four sib-pair linkage methods for analyzing sibships with more than two affecteds: interest of the binomial maximum likelihood approach. Genet Epidemiol 998; 5: 7 9. 9 Annesi-Maesano I. Rhinitis and asthma: epidemiological evidence. Allergy Clin Immunol Intern ; : 7. Lander E, Kruglyak L. Genetic dissection of complex traits: guidelines for interpreting and reporting linkage results. Nat Genet 995; : 4 47.

Sawcer S, Jones HB, Judge D et al. Empirical genomewide significance levels established by whole genome simulations. Genet Epidemiol 997; 4: 9. The Collaborative Study on the Genetics of Asthma (CSGA). A genome-wide search for asthma susceptibility loci in ethnically diverse populations. Nat Genet 997; 5: 89 9. Hizawa N, Freidhoff LR, Chiu YF et al. Genetic regulation of Dermatophagoides pteronyssinus-specific IgE responsiveness: a genome-wide multipoint linkage analysis in families recruited through asthmatic sibs. J Allergy Clin Immunol 998; : 46 44. 4 Mathias RA, Freidhoff LR, Blumenthal MN et al. Genomewide linkage analyses of total serum IgE using variance components analysis in asthmatic families. Genet Epidemiol ; : 4 55. 5 Koppelman GH, Stine OC, Xu J et al. Genome-wide search for atopy susceptibility genes in Dutch families with asthma. J Allergy Clin Immunol ; 9: 498 56. 6 Wjst M, Fischer G, Immervoll T et al. A genome-wide search for linkage to asthma. German Asthma Genetics Group. Genomics 999; 58: 8. 7 Howard TD, Postma DS, Hawkins GA et al. Fine mapping of an IgE-controlling gene on chromosome q: analysis of CTLA4 and CD8. J Allergy Clin Immunol ; : 74 75. 8 Hizawa N, Yamaguchi E, Jinushi E et al. Increased total serum IgE levels in patients with asthma and promoter polymorphisms at CTLA4 and FCERB. J Allergy Clin Immunol ; 8: 74 79. 9 Bradley M, Söderhäll C, Luthman H et al. Susceptibility loci for atopic dermatitis on chromosomes,, 5, 7 and 8 in a Swedish population. Hum Mol Genet ; : 59 548. Ober C, Cox NJ, Abney M et al. Genome-wide search for asthma susceptibility loci in a founder population. Hum Mol Genet 998; 7: 9 98. Blumenthal MN, Ober C, Beaty TH et al. Genome scan for loci linked to mite sensitivity: the Collaborative study on the genetics of asthma (CSGA). Genes Immun 4; 5: 6. Ober C, Tsalenko A, Willadsen S et al. Genome-wide screen for atopy susceptibility alleles in the Hutterites. Clin Exp Allergy 999; 9: S S5. Leynaert B, Neukirch C, Kony S et al. Association between asthma and rhinitis according to atopic sensitization in a population-based study. J Allergy Clin Immunol 4; : 86 9. 4 Kauffmann F, Dizier MH, Annesi-Maesano I et al. EGEA (Epidemiological study on the Genetics and Environment of Asthma, bronchial hyperresponsiveness and atopy) descriptive characteristics. Clin Exp Allergy 999; 9 (Suppl 4): 7. 5 Kauffmann F, Annesi-Maesano I, Liard R et al. Construction et validation d un questionnaire en épidémiologie respiratoire. L exemple du questionnaire de l étude épidémiologique des facteurs génétiques et environnementaux de l asthme, l hyperréactivité bronchique et de l atopie (EGEA). Rev Mal Resp ; 9:. 6 O Connell J, Weeks E. PedCheck: a program for identification of genotype incompatibilities in linkage analysis. Am J Hum Genet 998; 6: 59 66. 7 Satsangi J, Parkes M, Louis E et al. Two stage genome-wide search in inflammatory bowel disease provides evidence for susceptibility loci on chromosomes, 7 and. Nat Genet 996; 4: 99. 8 Majumder PP, Pal N. Nonrandom segregation: uniformly most powerful test and related considerations. Genet Epidemiol 987; 4: 77 87. 9 Alcais A, Philippi A, Abel L. Genetic model-free linkage analysis using the maximum-likelihood-binomial method for categorical traits. Genet Epidemiol 999; 7 (Suppl ): S467 S47. 4 Abel L, Muller-Myhsok B. Robustness and power of the maximum-likelihood-binomial and maximum-likelihoodscore methods, in multipoint linkage analysis of affectedsibship data. Am J Hum Genet 998; 6: 68 647. 4 Kruglyak L, Daly MJ, Reeve-Daly MP, Lander ES. Parametric and nonparametric linkage analysis: a unified multipoint approach. Am J Hum Genet 996; 58: 47 6.