|Home | About | Journals | Submit | Contact Us | Français|
Rationale: The basis for gender influences on allergen-specific IgEs is unclear.
Objectives: To perform regular and sex-stratified genomewide linkage analyses of IgE to each of three allergens (Ascaris lumbricoides, Blatella germanica [German cockroach]), and Dermatophagoides pteronyssinus [dust mite]) and to conduct an association study of a candidate gene in a linked genomic region.
Methods: Genomewide linkage analyses of allergen-specific IgEs were conducted in 653 members of eight large families of Costa Rican children with asthma. An analysis of the association between single-nucleotide polymorphisms in thymic stromal lymphopoietin (TSLP) and IgE measurements was conducted in 417 parent–child trios in Costa Rica. Significant results were replicated in 470 families of white children in the Childhood Asthma Management Program (CAMP).
Measurements and Main Results: Among all subjects, there was suggestive evidence of linkage (LOD 2.72) to IgE to Ascaris (on chromosome 7q) and IgE to dust mite (on chromosomes 7p and 12q). In a sex-stratified analysis, there was significant evidence of linkage to IgE to cockroach on chromosome 5q23 (peak LOD, 4.14 at 127 cM) in female subjects. TSLP is located within the 1.5 LOD-unit support interval for this linkage peak and has female-specific effects on lung disease in mice. In a sex-stratified analysis, the T allele of single-nucleotide polymorphism rs2289276 in TSLP was associated with reductions in IgE to cockroach (in Costa Rican girls) and total IgE (in girls in Costa Rica and in CAMP; P value for sex-by-genotype interaction, <0.01 in both studies).
Conclusions: Consistent with findings in murine models, a variant in TSLP may have female-specific effects on allergic phenotypes.
Genomewide linkage analyses of allergen-specific IgE measurements have been performed, but the results of sex-stratified analyses have not been presented.
This study identifies a female-specific locus for cockroach IgE. A SNP in a gene with female-specific effects on lung disease in mice (TSLP) is shown to be associated with cockroach IgE in Costa Rican girls and with total IgE in girls in two populations.
Childhood asthma is a major public health problem in nations with a Western lifestyle, such as Costa Rica (1), where most children with asthma are atopic (2, 3). A familial predisposition to produce serum immunoglobulin E (IgE) in response to allergen exposure is a hallmark of atopy (4, 5). Measurements of allergen-specific IgEs are important predictors of asthma and asthma morbidity in general and among Costa Ricans in particular (2, 3, 6, 7).
Allergen-specific IgE levels are independently heritable traits (5, 8). Gender is a major factor influencing allergen-specific IgE responses (9–11) and their heritability (12). Genomewide analyses that did not account for potential gender effects have failed to identify unique chromosomal loci with significant evidence of linkage to any allergen-specific IgE (13–17).
As is the case with allergen-specific IgEs, serum total IgE shows significant sexual dimorphism in humans (9–11). We conducted sex-stratified, genomewide linkage analyses and demonstrated distinct genetic architectures of total IgE among male and female subjects in Costa Rica (18). We hypothesized that, as demonstrated for total IgE, gender would play a prominent role in genomewide linkage analyses of allergen-specific IgE. In this study, we report results of regular and sex-stratified, genomewide linkage analyses of serum IgE to each of three common allergens (Ascaris lumbricoides [Ascaris], Blatella germanica [German cockroach], and Dermatophagoides pteronyssinus [dust mite]) in 653 members of eight extended families of Costa Rican children with asthma. We have identified a novel female-specific locus (chromosome 5q21-q32) for IgE to cockroach. Within this genomic region, we genotyped variants in a compelling candidate gene with experimental evidence of female-specific effects on lung disease (thymic stromal lymphopoietin [TSLP]) (19–23). We report an association between a single-nucleotide polymorphisms (SNP) in TSLP and IgE to cockroach in Costa Rican girls and total IgE in girls with asthma in Costa Rica and in the Childhood Asthma Management Program (CAMP). This work was partially presented in abstract form (24).
For details, see the online supplement.
The population of the Central Valley of Costa Rica is a genetic isolate (25, 26) of mixed Spanish and Amerindian ancestry with a prevalence of asthma that is among the highest in the world (1). For linkage analysis, eight large families of Costa Rican children with asthma were recruited as previously described (18). In brief, inclusion criteria for the eight probands included age at least 6 years but less than or equal to 14 years, physician-diagnosed asthma and more than two respiratory symptoms or asthma attacks in the previous year, increased airway responsiveness, more than one sibling with physician-diagnosed asthma, and at least six great-grandparents born in the Central Valley of Costa Rica.
The protocols for subject recruitment and data collection for parent–child trios have been previously described (3, 27). Children included in the study had asthma (defined as physician-diagnosed asthma and more than two respiratory symptoms or asthma attacks in the previous year) and high probability of having more than six great-grandparents born in the Central Valley (confirmed by our study genealogist in 94.8% of children [n = 416]). The latter criterion was required to increase the likelihood that children would be descendants of the founder population of the Central Valley (28). Of the 439 participating children, 426 had DNA that passed quality control and are included in this analysis along with their parents.
All study participants (with the exception of parents in trios, who only provided a blood sample) completed a protocol that included a questionnaire and collection of blood samples for DNA extraction and measurements of serum total and allergen-specific IgE. Levels of serum total and allergen-specific IgE were determined by the UniCAP 250 system (Pharmacia and Upjohn, Kalamazoo, MI), with samples measured in duplicate.
Written consent was obtained from adults. Written parental consent was obtained for children, for whom written assent was also obtained. The study was approved by the Institutional Review Boards of the Hospital Nacional de Niños (San José, Costa Rica) and Brigham and Women's Hospital (Boston, MA).
CAMP was a multicenter clinical trial of the effects of antiinflammatory medications in children with mild to moderate asthma. All recruited children had asthma defined by symptoms greater than two times per week, the use of an inhaled bronchodilator at least twice weekly or the use of daily medication for asthma, and airway responsiveness (29, 30). Children with severe asthma or other clinically significant conditions were excluded (29). Of the 1,041 children enrolled in the original clinical trial, 968 children and 1,518 of their parents contributed DNA samples. This analysis was restricted to 483 families of white (non-Hispanic) children.
Serum total IgE was measured by radioimmunosorbent assays from blood samples collected during the screening sessions of CAMP (29). Allergen-specific IgEs were not measured in CAMP.
CAMP was approved by the Institutional Review Boards of the Brigham and Women's Hospital and the other CAMP study centers.
The Genome Quebec Innovation Centre performed genomewide genotyping of 380 autosomal short-tandem repeat (STR) markers (average spacing, 8.2 cM) in 667 individuals with Applied Biosystems (Foster City, CA) 3700 and 3730 analyzers using DNA extracted from blood samples using Puregene kits (Gentra Systems, Minneapolis, MN). An additional nine STR markers on chromosome 7 were genotyped at Brigham and Women's Hospital as previously described (31).
RELPAIR (Version 2.0; University of Michigan, http://csg.sph.umich.edu/boehnke/relpair.php) was used to determine pedigree relationships on the basis of the genome scan marker data (32, 33). Mendelian inconsistencies at individual markers were resolved using PEDCHECK (34). Pedigree genotype inconsistencies were observed on average less than 0.5% per STR marker.
Using data from European Americans in the International HapMap project (35), we applied a linkage disequilibrium–tagging algorithm (minor allele frequency 10% and r2 0.8) to capture common variation in TSLP and its 10-kb flanks (36). SNPs were genotyped with the SEQUENOM iPLEX platform (Sequenom, San Diego, CA) (37). Of the seven SNPs that were successfully genotyped, one (rs11539838) was monomorphic. The remaining six SNPs capture at least 72% of the HapMap SNPs, with minor allele frequency of at least 10% in TSLP and its 10-kb flanks in European American trios at r2 0.8. One of these SNPs (rs2289276) was selected for replication studies in families of white children in CAMP and genotyped using the Taqman genotyping assay (Applied Biosystems) (38). In Costa Rica and CAMP, duplicate genotyping was performed on approximately 5% of the samples. Only one discordant genotype was detected, with genotype completion rates greater than 98.0% for all loci.
Estimates of narrow-sense heritability (h2N) and multipoint variance components linkage analysis for IgE to each of the allergens of interest were performed using SOLAR, version 2.12 (Southwest Foundation for Biomedical Research, San Antonio, TX) (39). Because values for each allergen-specific IgE were nonnormally distributed, the adjusted LOD score option in SOLAR was used for genomewide linkage analyses (40), which were conducted in all subjects and repeated after stratification by sex (18). Covariates considered in the linkage analyses included sex, age, ever-smoking, pack-years of cigarette smoking, age squared, and pack-years squared. Covariates that were significant (P < 0.05) were retained in the final models.
We selected one gene within the 1.5 LOD-unit support interval for the linkage peak to cockroach-specific IgE in female subjects for further analysis. TSLP was selected because it was the only gene within the genomic region of interest that had previously demonstrated a role in allergic disease in a sex-specific fashion (23). Association analyses were performed with the FBAT (Family-based Association Test) statistic implemented in PBAT version 5 (41). Sex-by-genotype interactions were tested using the FBAT-interaction statistic in PBAT (42). Analysis was limited to additive models. Measurements of IgE to cockroach were nonnormally distributed and rank-transformed before analysis. Given the nonmeasurement of IgE to cockroach in CAMP, we evaluated a correlated trait (total IgE, which was measured in both populations) in replication studies. In Costa Rica, cockroach-specific IgE is significantly correlated with (log10 transformed) total serum IgE (Spearman's correlation coefficient, 0.56; P < 0.0001). Total IgE levels were transformed to a log10 scale and adjusted for age (and sex where appropriate) for data analysis in Costa Rica and CAMP. For total IgE, we used P < 0.05 in both populations to define statistical significance in the setting of multiple tests (43).
Of the 667 members of extended pedigrees with genotypic data, allergen-specific IgE measurements were missing on 14 individuals. Table E1 in the online supplement shows the characteristics of the 653 members of the eight families included in the linkage analysis. There was significant variability among participating families with regard to number of members, gender, values of each allergen-specific IgE, and percentage of subjects with asthma.
Among all subjects, estimates of h2N of IgE to Ascaris, IgE to cockroach, and IgE to dust mite were 0.33 (SE, 0.072; P = 2.2 × 10−13), 0.39 (SE, 0.069; P = 1.6 × 10−17), and 0.38 (SE, 0.075; P = 3.6 × 10−13), respectively. Sex was the only covariate significantly associated with each of the IgEs. After stratification by sex, estimates of the h2N of IgE to Ascaris, IgE to cockroach, and IgE to dust mite among female subjects were 0.34 (SE, 0.12; P = 3.8 × 10−5), 0.52 (SE, 0.11; P = 3.8 × 10−10), and 0.28 (SE, 0.11; P = 9.6 × 10−5), respectively. Among male subjects, estimates of the h2N of IgE to Ascaris, IgE to cockroach, and IgE to dust mite were 0.47 (SE, 0.11; P = 8.2 × 10−9), 0.43 (SE, 0.11; P = 7.0 × 10−6), and 0.49 (SE, 0.13; P = 3.4 × 10−6), respectively. Thus, the estimated h2N of IgE to Ascaris and IgE to dust mite was lower in female subjects than in male subjects, but that of IgE to cockroach was higher in female subjects than in male subjects.
Table 1 and Figure 1 show the main results of the genomewide linkage analysis of serum allergen–specific IgEs in large families of Costa Rican children with asthma. Among all subjects, there was suggestive evidence of linkage (LOD 1.90) (44) to IgE to Ascaris on chromosome 7q (peak LOD, 2.23 at 145 cM), which was increased when the analysis was repeated after inclusion of nine additional STR markers (peak LOD, 2.72 at 153 cM). A sex-stratified analysis revealed only modest evidence of linkage to IgE to Ascaris in male and female subjects (peak LOD for male subjects, 1.96 on chromosome 7 at 54 cM; peak LOD for female subjects, 1.78 on chromosome 16 at 135 cM). Among all subjects, no genomic region showed at least suggestive evidence of linkage to IgE to cockroach (peak LOD, 1.83 on chromosome 17 at 25 cM). However, results of the sex-stratified linkage analysis of IgE to cockroach were markedly different between female and male subjects. Although no genomic region showed at least suggestive evidence of linkage in male subjects, there was significant linkage (LOD, 3.3) (44) to IgE to cockroach on chromosome 5 (peak LOD, 4.14 at 127 cM) in female subjects. The 1.5 LOD-unit support interval for this linkage peak encompasses chromosomal region 5q21-q32 and includes a candidate gene for atopy (TSLP). Three additional genomic regions showed suggestive evidence of linkage to cockroach-specific IgE (chromosomes 7q [LOD, 2.79 at 173 cM], 11q [LOD, 2.42 at 89 cM], and 17p [LOD, 3.24 at 28 cM]) in female subjects. Among all subjects, only one genomic region (chromosome 7p at 56 cM) showed suggestive evidence of linkage to IgE to dust mite. A sex-stratified analysis revealed only modest evidence of linkage to IgE to dust mite on chromosome 7p (peak LOD, 1.68 at 30 cM) in male subjects and chromosome 12q (peak LOD, 1.67 at 152 cM) in female subjects.
Baseline characteristics of children with asthma in Costa Rica and white (non-Hispanic) children with asthma in CAMP are presented in Table 2. Of the 426 participating parent–child trios in Costa Rica, 9 were excluded from this analysis because of Mendelian inconsistencies. Of the 483 white families in CAMP, 13 were removed from this analysis because of Mendelian inconsistencies, leaving 470 families and 503 probands. Parental genotypes were in Hardy-Weinberg equilibrium (P > 0.01) for all of the SNPs in Costa Rica and CAMP.
No SNP in TSLP was significantly associated with IgE to cockroach in all subjects in Costa Rica. However, the T allele of a SNP in the 5′ untranslated region of TSLP (rs2289276) was inversely associated with IgE to cockroach in Costa Rican girls (P = 0.03). Although this is not statistically significant, there was a positive association between the T allele of rs2289276 and IgE to cockroach in Costa Rican boys (P = 0.18). Although there was no significant evidence of a sex-by-genotype interaction for SNP rs2289276 on IgE to cockroach (P = 0.19), there was limited statistical power to detect such an interaction.
Consistent with the findings for IgE to cockroach, no SNP was significantly associated with total IgE among all Costa Rican children. However, a sex-by-SNP (rs2289276) interaction on total IgE was statistically significant in Costa Rica and CAMP (P < 0.01 in both studies). The T allele of SNP rs2289276 was inversely associated with total IgE in Costa Rican girls (P = 0.01) and in girls in CAMP (p = 0.004). There was no statistically significant association between the T allele of SNP rs2289276 and total IgE among boys in Costa Rica or CAMP (Table 3).
This is the first report of a genomewide, sex-stratified linkage analysis of serum allergen–specific IgEs in families of children with asthma. Although sex had no strong influence in our findings for IgE to Ascaris and IgE to dust mite, we found distinct genetic architectures for cockroach-specific IgE in male and female subjects and identified significant linkage to IgE to cockroach on chromosome 5 (peak LOD, 4.14 at 127 cM) in female subjects. The T allele of a SNP (rs2289276) in a candidate gene near this linkage peak (TSLP) was inversely associated with cockroach-specific IgE in Costa Rican girls. In addition, this allele was inversely associated with total IgE in girls in each of two family-based study samples (in Costa Rica and in CAMP). To our knowledge, this is the first sex-specific association with allergy-related phenotypes to be replicated in independent populations (45).
Among members of the eight large families of Costa Rican children with asthma, the estimated heritability of IgE to Ascaris, IgE to dust mite, and IgE to cockroach was similar in all subjects (h2N range, 0.33–0.39) and in male subjects only (h2N range, 0.43–0.49). In contrast, the estimated heritability of IgE to cockroach was higher (h2N, 0.52) than that of Ascaris-specific IgE (h2N, 0.34) or dust mite–specific IgE (h2N, 0.28) among female subjects, suggesting that genetic factors play a more significant role in the production of IgE to cockroach than in the production of IgE to Ascaris or dust mite in female Costa Ricans.
We report the first genomewide linkage analysis of IgE to Ascaris (a phenotype associated with asthma severity in Costa Rican children) (3). Among all members of large Costa Rican families, there was suggestive evidence of linkage to Ascaris-specific IgE on chromosome 7q, a genomic region previously linked to obesity (a trait often associated with asthma) (46, 47). Previous genomewide studies in families of subjects with asthma revealed two chromosomal regions (19p13 [LOD, 3.51] and 20q13 [LOD, 4.93]), with significant evidence of linkage to dust mite allergy (assessed by skin test positivity and not by continuous measurement of specific IgE) (48, 49). Although “usual” genomewide linkage analyses of allergen-specific IgE measurements have not revealed loci meeting criteria for genomewide significance, an analysis that took maternal imprinting into account showed significant evidence of linkage to dust mite–specific IgE on chromosome 8p22 (maximized LOD score with respect to disease-model parameters, 4.76) (50). There was little to no overlap in our findings for each allergen-specific IgE, suggesting that these phenotypes are influenced by different genomic regions.
The 1.5 LOD-unit support interval for the observed linkage peak to cockroach-specific IgE in female subjects is broad (chromosome 5q21-q32, encompassing ~87 known genes) and includes a region previously shown to be significantly linked to mite-sensitive asthma in Japanese subjects (17). Although genes within 5q21-q32 (e.g., IL4, IL13, and CD14) (52–54) have been previously examined, this is the first association study of TSLP for asthma- and allergy-related phenotypes.
Gender-specific analyses of genetic association can lead to false-positive results due to multiple testing (45). Studies that claim to identify sex-specific effects are often limited by lack of replication in independent populations, absence of replication of the findings with regard to the direction of the original association, and no formal testing for sex-by-genotype interaction. We have tested for sex-by-genotype interaction and replicated our female-specific association between TSLP and total IgE in two independent populations for the same SNP, in a consistent direction of association.
TSLP, a gene with two splice variants, can prime dendritic cells and stimulate production of Th2 cytokines by naive T cells (54), and TSLP knockout mice are protected from allergic airway responsiveness (55). Increased expression of TSLP in the airways of human subjects with asthma is associated with increased levels of Th2 cytokines and airway responsiveness (56). Of relevance to our findings, overexpression of TSLP in transgenic mice leads to prominent pulmonary pathology, including perivascular leukocytic infiltration and prominent eosinophilia, with more severe effects in female mice than in male mice (23).
Our study has several limitations. First, we had strong motivation to perform sex-stratified, genomewide linkage analyses of allergen-specific IgEs (9–11, 18) and found evidence for an inverse association between the T allele of SNP rs2289276 in TSLP and IgE to cockroach in Costa Rican girls, but replication of this finding is needed. However, we are encouraged by our finding that this allele was also inversely associated with total IgE in Costa Rican girls and girls in CAMP. Second, evidence of a sex-by-SNP (rs2289276) interaction for total IgE in Costa Rica and CAMP may implicate that TSLP is involved more generally in the sex-dependent regulation of total and allergen-specific IgEs. Third, the functional effects of TSLP polymorphisms are not known. However, SNP rs2289276 is predicted to disrupt an exonic splicing enhancer site (57). Finally, data from the International HapMap consortium suggest that SNP rs2289276 is within a large block of linkage disequilibrium that includes WDR36 (WD repeat-containing protein 36), a gene involved in T-cell activation. We cannot rule out the possibility that other SNPs within WDR36 alternately or additionally explain our findings of association. Although our findings provide evidence to support a role of TSLP in human allergy (in a sex-specific fashion), a fine-mapping association study of chromosome 5q21-q32 is needed to firmly identify the gene(s) responsible for the observed female-specific linkage peak to IgE to cockroach.
In summary, we have identified a female-specific locus for cockroach-specific IgE on chromosome 5q21-q32 in Costa Ricans and a female-specific association between a variant in a candidate gene in this region (TSLP) and IgE to cockroach in Costa Rica and total serum IgE in Costa Rica and CAMP. The strong influence of gender on age-associated asthma prevalence (58) and expression of various atopic phenotypes has been well documented but largely unexplored. Our results suggest an important role of sex-specific genetic effects on allergic phenotypes.
The authors thank all families for their invaluable participation in the Genetics of Asthma in Costa Rica and the CAMP studies. The authors acknowledge the CAMP investigators and research team for their help in data collection and the CAMP investigators and research team, supported by NHLBI, for collection of CAMP Genetic Ancillary Study data. All work on data collected from the Genetics of Asthma in Costa Rica and the CAMP Genetic Ancillary Study was conducted at the Channing Laboratory of the Brigham and Women's Hospital under appropriate CAMP policies and human subject's protections. The authors also thank Mr. John Ziniti for help in the preparation of this manuscript.
The Genetics of Asthma in Costa Rica Study is supported by NIH grants HL66289, HL04370, and HL074193. The CAMP Genetics Ancillary Study is supported by the NHLBI, N01-HR-16049. Additional support for this research came from grants U01HL065899 and P50HL67664. G.M.H. is the recipient of an Individual National Research Service Award (1F32HL083634).
This article has an online supplement, which is accessible from this issue's table of contents at www.atsjournals.org
Originally Published in Press as DOI: 10.1164/rccm.200711-1697OC on January 31, 2008
Conflict of Interest Statement: None of the authors has a financial relationship with a commercial entity that has an interest in the subject of this manuscript.