A Nonsynonymous FCER1B SNP is Associated with Risk of Developing Allergic Rhinitis and with IgE Levels

Allergic rhinitis is associated with elevated serum IgE levels. IgE response is mediated by the high-affinity IgE receptor (FcεRI), which is polymorphic. Studies analyzing the association between allergic rhinitis and FcεRI variants have been conducted with controversial results. The objective of this study is to analyze, in 1,041 individuals, the putative clinical association of allergic rhinitis with common polymorphisms in FcεRI subunits genes. These SNPs included FECR1A rs2494262, rs2427837 and rs2251746; FECR1B rs1441586, rs569108 and rs512555; FCER1G rs11587213, rs2070901 and rs11421. Statistically significant differences were observed for the FCER1B rs569108 and rs512555 polymorphisms frequencies when comparing patients with allergic rhinitis without asthma and controls. The OR (95% CI) value for the 237Gly allele (rs569108) is equal to 0.26 (0.08–0.86, P = 0.017) and for the G allele (rs512555) it is equal to 0.27 (0.08–0.88, P = 0.020). These two SNPs are linked (D’ = 1.0, LOD = 56.05). Also observed was a statistically significant trend towards lower IgE values among allergic rhinitis patients with variant alleles for both SNPs. In conclusion, in patients with allergic rhinitis without asthma, the FCER1B rs569108 and rs512555 polymorphisms are associated with increased risk of developing allergic rhinitis and with lower IgE levels.


Results
The SNP frequencies observed in the study group are consistent with those previously described in the 1000 genomes public database for individuals with European descent 24 , and are at Hardy-Weinberg equilibrium. The genotype and allele frequencies for the nine SNPs analyzed in overall allergic rhinitis patients and control individuals are summarized in Table 1. No major frequency differences were observed when comparing patients and controls.
Patients were stratified in two subgroups according to their clinical presentation; allergic rhinitis alone and allergic rhinitis + asthma. Table 2 shows the FCER1 genotypes. Among patients with allergic rhinitis alone the frequencies for individuals carrying the minor alleles for the SNPs rs569108 and rs512555 were remarkably reduced among allergic rhinitis patients as compared to the control individuals. In fact, only three allergic rhinitis patients (representing 2% of this clinical subgroup) carried such minor alleles in heterozygosity, whereas 37 controls (7.2%) carried such variant alleles. It is to be noted that no carriers of these variant alleles in homozygosity were observed among all patients and controls (Table 1). Statistical significance was also observed in the test for trend for the SNPs rs569108 and rs512555 (χ 2 = 5.86, P = 0.015; χ 2 = 5.60, P = 0.018, respectively). The SNP rs569108 corresponds to a missense variant E237G, which is predicted as deleterious by Sorting Intolerant From Tolerant (SIFT) 26 , and predicted as possibly damaging by the Polymorphism Phenotyping v2 (PolyPhen-2) prediction 27 . The SNP rs512555 is a downstream gene variant whose functional effect is unknown; the high linkage between these two SNPs in the present study (D' = 1.0, LOD = 56.05) suggests that the genetic association observed in Table 2 might actually be related to the missense SNP rs569108. The rest of the SNPs analyzed did not reveal any differences either in genotypes or in allele frequencies, when patients with allergic rhinitis alone and control individuals were compared (Table 2). Among patients with allergic rhinitis + asthma none of the SNPs analyzed revealed significant differences when compared to control individuals (Table 2).
Because it has been suggested that FCER1 genotypes may modify IgE response 21 , we determined IgE values and analyzed their distribution according FCER1 genotypes. These findings are summarized in Table 3. A statistically significant trend can be observed towards lower IgE values among patients with variant alleles for the SNPs rs569108 and rs512555, confined to patients with allergic rhinitis alone. Among patients with allergic rhinitis + asthma who carried the above-mentioned variant alleles, the IgE values were lower as compared to non-carriers, but the difference was not statistically significant. The rest of the variant alleles were not related to the IgE levels. Since a statistical significant association of the FCER1B variant alleles with IgE levels in patients with allergic rhinitis alone was detected, linear regression under the standard additive model, including in a single model all genotypes, gender, age, familial antecedents of allergic diseases, smoking and prick tests, was studied with regard to the IgE values. Only age, with lower IgE concentrations in older age, (p = 0.019) and the mutated genotypes corresponding to the FCER1B SNPs rs569108 and rs512555 genotypes (p = 0.005) remained significant. The rest of the putative genetic associations were discarded in the multiple analysis. No associations of genotypes or IgE levels were identified when individuals were stratified with regard to allergic rhinitis type (intermittent, persistent mild and persistent moderate), eosinophilia, FEV or asthma severity step.

Discussion
Allergic rhinitis, like other complex diseases, results from interplay of genetic and environmental risk factors [3][4][5] . The heritability factor has been calculated to be quite high, between 74-82% 1 and 90% 2 , and several genetic variants are known to be associated with allergic rhinitis risk in some populations 3,28,29 . However, associations have been described with genes which have a weak, or unknown, relation with allergic rhinitis pathogenesis, and it has been speculated that the putative associations could be due to hitherto unknown epigenetic mechanisms 3 . In contrast, the present study focuses on genes that are functionally related to the pathogenesis and the clinical presentation of this disease.
Our results show association of two SNPs in the FCER1B gene with the risk of developing allergic rhinitis: the nonsynonymous Glu237Gly (rs569108) and the 3′ UTR SNP (rs512555). The Glu237 allele is more common among allergic rhinitis alone patients than in healthy subjects. When allergic rhinitis + asthma patients were compared to healthy subjects, we observed no statistically significant differences in the allele frequencies or genotypes. The FCER1B Glu237Gly variant has been studied in a Japanese population (233 patients with nasal allergy and 100 control subjects) showing an association between Gly 237 and nasal allergy and atopy 30 . Recently, a meta-analysis including these two polymorphisms 16 has been published. Such study includes 4496 asthmatics and 4571 controls for the E237G variant and 2005 cases and 1868 controls for rs1441586 from several populations: Caucasians, Africans and Asians. The results obtained suggest that the 237Gly allele might influence the risk for developing atopic asthma, although no significant association was observed in the total population. Regarding the rs1441586 SNP, a significant association with asthma risk was observed among Asians only 16 . Our findings do not support such an association with asthma, but they do support association with rhinitis. To date, no meta-analyses with allergic rhinitis patients have been performed. The close relationship between asthma and allergic rhinitis is well known. A majority of adult patients with asthma also suffer from allergic rhinitis 17 and vice-versa (this study). However, the presence or absence of allergic rhinitis + asthma can be an important confounding factor in the studies carried out so far, which did not discriminate these clinical presentations. In this sense careful phenotyping using strictly defined guidelines should be used to elucidate whether the putative associations with FECR1 polymorphisms are related to asthma or to concomitant allergic rhinitis. In this sense, we cannot rule out that environmental factors may modify the observed associations. Further PheWas studies would be necessary to elucidate this point. The functional mechanism underlying the association found in our study is a matter of speculation, and there are certain aspects which should be taken into consideration. The regulation of the expression of Fcε RI is crucial to the receptor function. It is known that both the expression of Fcε RIβ subunit and the serum levels of IgE are important regulating factors. Fcε RIβ regulates the receptor response to IgE through its capacity to amplifying receptor signaling. IgE-Fcε RI binding causes overexpression of the receptor 18,20 with an increase in the total Fcε RIα content on the surface. This process can be reverted by the removal of IgE, which results in a reduction in receptor expression and Fcε RIα content 19 . The relation between FCER1B variants and IgE levels is controversial. A wide variation in the IgE levels, consistent with previous studies, was observed in the present study 31,32 . We observed that the FCER1B variants are associated with lower serum IgE levels, in a group of patients phenotyped to discriminate between patients with allergic rhinitis without asthma and allergic rhinitis + asthma. The association is lost when all allergic rhinitis patients are merged as a single clinical group, as has been done in most previous studies. Another source of variability with previous findings is that the variant alleles analyzed in this study show different frequencies in diverse ethnicities 24 ; these differences could explain, at least in part, the different results described to date.
Several studies have reported an association between several FCER1A variants and total serum IgE levels 21 as well as increased risk for atopic eczema and asthma. We failed to replicate the previously described association of IgE levels with the FCER1A polymorphism rs2427837, although it is important to note that this association was described in a specific population corresponding to a general population of individuals living in Augsburg 21 .  Another study in a Han Chinese population failed to find an association between FCER1A variants and IgE level or allergic allergic rhinitis susceptibility 22 . Summing up, we have identified an association of FCER1B variants with the risk of developing allergic rhinitis alone, and with IgE levels in these patients. Further association studies using well-phenotyped patients are required to identify genetic associations that, so far, may have been overlooked in studies considering heterogeneous groups of patients.

Methods
Study Population. We studied a cohort of 1,041 individuals, composed of 515 unrelated patients with allergic rhinitis and 526 unrelated healthy controls. Written consent for participation was obtained for all participants. All the patients and 97% of the control subjects who were invited to participate in the study agreed to do so.
Of the patients, 265 were recruited from the Allergy Department, Infanta Cristina Hospital (Badajoz, Spain) and 250 were recruited from the UGC Allergy, Regional Hospital (Málaga, Spain). Clinical characteristics of the participants are summarized in Table 4. Most of these patients had participated in a previous study and their clinical characterization has been described elsewhere 33 . Diagnosis for allergic rhinitis was based on the diagnostic criteria defined by the Joint Task Force on Practice Parameters in Allergy, Asthma and Immunology 34 . All patients were diagnosed by an allergy specialist and had a positive skin prick test 35 for one or more aeroallergens. Of the 515 patients with allergic rhinitis, 366 (71.1%) was diagnosed with allergic asthma, according to American Thoracic Society criteria 36 . These patients were classified as defined by the step classification of the Global Initiative for Asthma guidelines 37 . A total of 526 unrelated healthy controls, from the same geographic area, and ethnically matched with patients, were included in the study. Medical history was obtained and examination was performed for each participant to exclude pre-existing disorders. Individuals with familial (up to second-degree relatives) or personal antecedents of allergic, atopic or autoimmune diseases were excluded. Control subjects were selected from staff and from medical students belonging to the Hospitals and the Universities participating in the study.
No statistically significant differences between cases and controls were observed for gender. Age was not different when comparing the overall group of patients and controls (T-Test P = 0.190). However patients with allergic rhinitis alone had a mean age slightly lower than controls (P = 0.035). The percentage of smokers was lower among patients with rhinitis plus asthma (Chi-Square test P = 0.002) as compared to control individuals, whereas the percentage of smokers was higher among patients with allergic rhinitis alone (P = 0.002) as compared to control individuals. The protocol for this study was in accordance with the Declaration of Helsinki and its subsequent revisions, and it was approved by the respective Ethics Committees of the participating Hospitals. All individuals were Spanish Caucasians. Genotype analysis. Genomic DNA was obtained from leukocytes and purified according to standard procedures. The selection of the SNPs analyzed was based on allele frequencies in the study population (higher than 0.01), as well as functional or clinical relevance, in accordance with the public 1000 genomes database 24 and published evidence 38,39 . SNP analyses were performed by means of TaqMan assays (Life Technologies, Alcobendas, Spain). Details of these assays are provided in supplemental Table S1. Detection was carried out by qPCR in a 7500 Real Time qPCR system (Life Technologies, Alcobendas, Spain). The amplification conditions were those recommended for the manufacturer. All SNP analyses were carried out in triplicate.

Statistical analyses.
Statistical analyses were carried out by using the Statistical Package for the Social Sciences (SPSS) version 19.0 for Windows (SPSS Inc., Chicago, Ill. USA). Genotype, allelic and Hardy-Weinberg equilibrium analyses were performed by using the software PLINK 40 .
The comparison of the frequencies for alleles and genotypes between patients and control individuals was performed with the χ 2 statistics or Fisher exact test, as appropriate. The putative association between the SNPs and the risk of developing rhinitis was estimated by determining the odds ratio (OR) and the 95% confidence interval (CI). Multiple comparison adjustments were done by using the False Discover Rate (FDR) correction 41 .  The statistical significance of the differences in analytical or clinical parameters were calculated by using the non-parametric Kruskall-Wallis test for independent samples. The gene-dose effect was analyzed by using the test for trend (Cochran-Armitage test). Potential confounders considered in this study were the type of allergic rhinitis, age, gender, familial antecedents of the disease, smoking habits, the presence of asthma and asthma severity step. Multiple analyses were performed by using logistic regression. Differences in the parameters determined were considered to be statistically significant when adjusted p values were lower than 0.05. In order to calculate the statistical power of the study, we evaluated the sample size with a genetic model based on the frequency for carriers of the risk SNP with an OR value = 2 (α = 0.05). According to the sample size in this study, the statistical power (two-sided associations) for the risk related to the presence of the variant alleles was higher than 99.8% for all SNPs except for the SNPs rs569108 (power = 93.2%) and rs512555 (power = 92.6%).