Association of IL1A, IL1B, and TNF Gene Polymorphisms With Chronic Rhinosinusitis With and Without Nasal Polyposis: A Replication Study

Objective  To replicate and extend recent findings in a Turkish population of associations between chronic rhinosinusitis (CRS) with nasal polyposis and single-nucleotide polymorphisms (SNPs) in the IL1A (rs17561 and Ser114Ala), IL1B (rs16944), and TNF (rs361525 and rs1800629) genes.

Design  In a case-control replication study, DNA samples were obtained from 206 patients with severe CRS (cases) and from 196 postal code–matched controls. For IL1A and TNF, the 3 reported SNPs were complemented with tagging SNPs using an International HapMap genotyping data set to ensure complete genetic coverage. For IL1B, only the single reported SNP was assessed. A total of 24 SNPs (7 in IL1A, 1 in IL1B, and 16 in TNF) were individually genotyped. The PLINK software package was used to perform genetic association tests.

Setting  Academic research.

Patients  Canadian population of individuals with severe CRS.

Main Outcome Measures  Allelic differences between cases and controls.

Results  Significant allelic differences between cases and controls were obtained for IL1A rs17561 (odds ratio [OR], 1.48; P = .02). The following 3 additional SNPs in this gene were associated with CRS: rs2856838 (OR, 0.63; P = .003), rs2048874 (OR, 0.57; P = .01), and rs1800587 (OR, 1.49; P = .02). These 3 SNPs remained significant after correction for multiple testing. No association was found with IL1B or TNF.

Conclusions  We replicated the previously reported association between the IL1A polymorphism and severe CRS and identified 3 potential new associations in the same gene. This further supports the potential contribution of IL1A to the development of CRS. We were unable to replicate previous reports of associations with IL1B or TNF.

Genetic factors are believed to have an important role in many common complex disorders. This fact, together with the identification of many single-nucleotide polymorphisms (SNPs) throughout the genome and the rapidly falling costs of genotyping, has contributed to the proliferation of association studies in epidemiologic genetics.^1,2 Replication of results of a genetic disease association study in independent samples has emerged as a standard for demonstrating the relevance of a candidate gene for a complex trait.³

Chronic rhinosinusitis (CRS) is a common inflammatory disorder involving the sinus mucosa. Patients with CRS report low quality-of-life index values for domains of bodily pain and social functioning.⁴ Biopsy specimens obtained at the time of surgery demonstrate an inflammatory process that is also colonized with nasal and exogenous bacteria, which are believed to contribute to the disease process.⁴ Chronic rhinosinusitis is further subclassified according to the presence or absence of nasal polyposis. Nasal polyposis is characterized by proliferation of the epithelial layer, glandular hyperplasia, thickening of the basal membrane, edema, focal fibrosis, and cellular infiltration of the stromal layer. In addition, the inflamed mucosa shows an accumulation of inflammatory cells, with production of numerous proinflammatory cytokines.⁴ Various factors are thought to influence the severity of inflammatory disorders, and it is hypothesized that there is an important genetic component. Genetic factors may affect cytokine gene expression, with repercussions in the severity of the inflammatory process.^5-7

Interleukin 1 (IL-1) is a pivotal cytokine involved in most inflammatory responses. Interleukin 1 is regularly expressed in nasal polyps, including epithelial cells and macrophages.^7,8 Interleukin 1 activates T cells and monocytes and upregulates expression of adhesion molecules. Interleukin 1 also induces expression of numerous cytokines and inflammation-associated proteins, which modulate the cascade of inflammatory responses. In humans, IL-1 exists in 2 forms, IL-1α (IL1A gene [OMIM 147760]) and IL-1β (IL1B gene [OMIM 147720]), located on chromosome 2 in both forms.⁵ Several studies have demonstrated that polymorphisms in IL1A are associated with atopy in adults without asthma,⁹ with nasal polyposis in adults with asthma,¹⁰ and with periodontitis.¹¹ Polymorphisms in IL1B have been associated with gastric cancer¹² and with inflammatory bowel disease.¹³

Tumor necrosis factor (TNF) is a crucial proinflammatory cytokine secreted predominantly by monocytes, macrophages, and T cells.⁷ The TNF gene (TNF; OMIMg 191160) is located within the highly polymorphic major histocompatibility complex region of chromosome 6. It exerts a range of inflammatory and immunomodulatory activities that are important in host defense.⁷TNF has been putatively implicated in the pathogenesis of diverse disease states, including increased susceptibility to infections, autoimmune disorders, neoplasia, neurodegenerative diseases, and even drug dependencies.⁷ Polymorphisms of TNF have been associated with asthma.¹⁴

A 2007 study⁶ identified IL1A (−4845GT and −4845TT [rs17561]]), IL1B (−511CC [ rs16944]]), and TNF (−238AA [rs361525] and −308GA [rs1800629]) as genotypes that are associated with nasal polyp susceptibility in a cohort of 82 Turkish patients with nasal polyposis. In this study, we aimed to replicate the CRS associations recorded for IL1A, IL1B, and TNF in a cohort of Canadian patients with severe CRS. We further aimed to extend on these findings by assessing associations across the entire genes for IL1A and TNF.

A total of 206 patients with severe CRS (with and without nasal polyposis) (hereinafter referred to as cases) and 196 controls were recruited prospectively. According to 2004 American Academy of Otolaryngology–Head and Neck Surgery guidelines,¹⁵ severe CRS was defined as the following: (1) persistent signs or symptoms of CRS despite previous endoscopic sinus surgery or (2) a history of more than 1 endoscopic sinus surgery procedure for CRS, regardless of outcome. A standardized questionnaire was administered assessing age, sex, race/ethnicity, smoking, seasonal and perennial allergies, physician-diagnosed asthma, and acetylsalicylic acid intolerance. Information was recorded about disease-related factors, including age at diagnosis, age at first sinus surgery, number of previous surgical procedures, medications required for management of the disease, and assessment of whether the disease was controlled with medication. Initial diagnoses of CRS with or without nasal polyposis were classified. All cases with early-onset nasal polyposis had previously undergone sweat chloride testing to rule out a diagnosis of cystic fibrosis.

Blood samples were collected (BD Vacutainer Serum Separator Tubes; BD Diagnostics, Franklin Lakes, New Jersey) and stored at 4°C until analysis for DNA extraction. Total IgE measurements were performed (DPC Immulite System; Diagnostic Products Corporation, Siemens, Los Angeles, California).

Controls were recruited from the following: (1) spouses or nonblood relatives living in the same household as the case or (2) individuals recruited by random telephone screening matched to the case's postal code. To minimize differences secondary to potential environmental exposures, the only attempt at matching cases and controls was their geographic location. Nevertheless, a standardized questionnaire assessing age, sex, and race/ethnicity was obtained for controls. A kit (Oragene; DNA Genotek, Ottawa, Ontario, Canada) was used for saliva collection and was sent to controls with prepaid return postage. As recommended by the manufacturer, saliva samples were stored at room temperature until genotyping.

The study was approved by McGill University Health Centre and the Centre Hospitalier de l’Université de Montréal Hôtel-Dieu surgical ethics committees. All cases and controls provided signed informed consent.

DNA was isolated from peripheral blood leukocytes. Blood was collected in citrate-treated tubes, and DNA was isolated using a kit (Puregene DNA; Gentra Systems, Germantown, Maryland) according to the high-throughput protocol for 10 mL of whole blood provided with the kit. DNA obtained from saliva was purified per the manufacturer's protocol (DNA Genotek). Isolated DNA from blood and saliva was stored at −80°C before use.

To ensure complete coverage of IL1A and TNF, a maximally informative set of SNPs was selected using the Centre d’Etude du Polymorphisme Humain genotype data from the International HapMap project¹⁶ covering 10 kilobases (kb) upstream and downstream for both genes. From this data set, a set of tagging SNPs was selected for each gene using a pairwise tagging algorithm implemented in an available software program (Haploview, version 3.2; http://www.broadinstitute.org/mpg/haploview).¹⁷ Minor allele frequency and r² thresholds were set at 0.05 and 0.8, respectively. IL1A rs17561 previously identified by Erbek et al⁶ was force included. For IL1B, genotyping was limited to rs16944 (previously reported⁶). Overall, we genotyped 7, 1, and 16 SNPs in IL1A, IL1B, and TNF, respectively, for a total of 24 SNPs.

Single-nucleotide polymorphisms were genotyped using a matrix-assisted laser desorption ionization–time-of-flight mass array spectrometer (Sequenom, San Diego, California). Primers were designed using available software (SNP Assay Design, version 3.0 for iPLEX reactions; Sequenom). The protocol and the reaction condition were in accord with the manufacturer's instructions.

Markers were excluded if they deviated significantly from Hardy-Weinberg equilibrium (P < .01 in controls), if they had low minimum allele frequency (<0.05), or if they had a call rate of less than 90% in cases and controls combined. The term call rate is an indication of the percentage of success of genotyping for a particular SNP and represents the percentage of samples that were successfully genotyped. Odds ratios (ORs) and 95% confidence intervals (95% CIs) were calculated. Allele, genotype, and haplotype frequencies were compared between cases and controls. Association analysis was performed by comparing allele frequencies between cases and controls using the χ² test. To correct for multiple testing, we used a method described by Nyholt.¹⁸ Logistic regression models were performed using sex as a covariate to calculate ORs for homozygous and heterozygous genotypes. All association tests and the logistic regression analysis were performed using available software (PLINK, version 1.02; http://pngu.mgh.harvard.edu/~purcell/plink/).¹⁹ The linkage disequilibrium plots were visualized using available software (Haploview, version 3.2).

Sample size was designed to provide 95% power to detect common alleles (>10%) that confer a 3.0-fold increase in risk. It was also designed to provide 50% power to detect common alleles (>25%) that confer a 2.0-fold increase in risk.

The clinical characteristics of the study population are given in Table 1. The mean (SD) ages of cases and controls were 52.3 (13.0) years and 48.8 (15.0) years, respectively. For cases, the initial diagnosis was mainly CRS with nasal polyposis (74.8%). The mean number of previous surgical procedures was 3.2, with a mean age at first sinus surgery of 38.1 years. History of atopy and history of asthma were present in 65.5% and 63.7%, respectively. Current smoking was present in 11.2%. Measured serum biomarkers showed median circulating eosinophilia of 3.6%, with 33.5% of cases demonstrating more than 5% eosinophilia. The median total serum IgE level was 0.087 μg/L (to convert IgE level to milligrams per liter, multiply by 0.001), with 41.7% having IgE levels of at least 0.12 μg/L.

All SNPs in IL1A, IL1B, and TNF were successfully genotyped (Table 2). Only the SNPs meeting the quality control and SNPs with a minimum allele frequency of 0.05 or higher and Hardy-Weinberg equilibrium of P ≥ .01 were considered for genetic association tests (Table 3).

An association with CRS was noted for the following 4 SNPs in IL1A: the previously reported⁶ rs17561 (OR, 1.48; P = .02) and 3 other SNPs (rs2856838 [OR, 0.63; P = .003], rs2048874 [OR, 0.57; P = .01], and rs1800587 [OR, 1.49; P = .02]). Only 3 SNPs (rs2856838, rs2048874, and rs1800587) remained significant after Nyholt correction for multiple testing (P ≤ .02). However, for rs17561, we have replicated results for the TT homozygote genotype (OR, 3.39; P = .007). The protective effect of rs2856838 (OR, 0.38; P = .002) and the risk effect of rs1800587 (OR = 3.16, P = .008) are enhanced with the homozygote form of the minor allele. In contrast, no association was found with SNPs in IL1B or TNF (Table 3).

Adjustment for sex as a covariate among the 4 significant SNPs in IL1A showed no difference in the risk of CRS. Assessment of association of these SNPs in the population showed no increase in the strength of the association for the subgroups with nasal polyposis or asthma, confirming that the observed relationship is not secondary to underlying asthma and is not limited to the subgroup having CRS with nasal polyposis.

The linkage disequilibrium pattern for IL1A in our population is shown in the Figure. Strong linkage disequilibrium is noted between rs17561 and rs1800587 (r² = 0.96), indicating that the SNPs represent the same association signal. Two haplotypes that included these 2 SNPs and whose frequencies in cases were statistically different from those in controls are GTC (P = .002) and TCT (P = .01).

Our objective was to replicate results from a previous study⁶ showing association between IL1A, IL1B, and TNF polymorphisms and CRS with nasal polyposis. In this study, we confirm the association between IL1A rs17561 and CRS and extend these results by identifying 3 additional IL1A SNPs associated with CRS. However, we did not replicate previously reported⁶ associations of IL1B and TNF polymorphisms with CRS.

Although we have replicated the association, the means by which the rs17561 polymorphism contributes to the development of disease remains unexplained. Because rs17561 represents a nonsynonymous mutation (Ser114Ala), this may lead to an altered protein with a potential functional effect in CRS.

We also report the following 3 new SNPs not previously associated with CRS: rs1800587, rs2048874, and rs2856838. Located in the promoter, rs1800587 is in tight linkage disequilibrium with rs17561. In a Brazilian population, rs1800587 has also been associated with chronic periodontal disease.¹¹ To the best of our knowledge, IL1A rs2048874 and rs2856838 have not previously been associated with CRS or other diseases.

In this study, IL1B rs16944 was not associated with severe CRS or nasal polyposis. A lack of association of this SNP in IL1B was reported in another study¹⁰ conducted among a Finnish population with nasal polyposis and asthma.

Although Erbek et al⁶ showed that TNF (−238 [rs361525] and −308 [rs1800629]) was associated with susceptibility to nasal polyposis, TNF polymorphisms were not associated with CRS in our study. The AA genotype for rs361525 associated with CRS in the Turkish population⁶ was not found in our population.

TNF has been associated with several inflammatory diseases, including asthma,¹⁴ atopy,²⁰ and chronic obstructive pulmonary disease.²¹ However, these results have not been confirmed in other studies.^22-25 The lack of reproducibility may be ascribed to small sample sizes, biologic and phenotypic complexity, population-specific linkage disequilibrium, effect size bias, or population stratification.³

In conclusion, our data confirm a previous study⁶ implicating an IL1A polymorphism and nasal polyposis. We identify 3 additional IL1A SNPs associated with CRS. One of them (rs1800587) is in tight linkage disequilibrium with the previously reported SNP (rs17561). No association with the disease was observed for SNPs in IL1B or TNF.

Understanding the role of IL1A will take us a step further in our understanding of the pathogenesis of CRS and should guide us toward more effective treatment and screening for this inflammatory disease.

Correspondence: Martin Desrosiers, MD, Department of Otolaryngology, Centre de Recherche du Centre Hospitalier de l’Université de Montréal Hôtel-Dieu, 3840 Rue St-Urbain, Montreal, QB H2W 1T8, Canada (desrosiers_martin@hotmail.com).

Submitted for Publication: April 20, 2009; final revision received July 17, 2009; accepted September 1, 2009.

Author Contributions: Ms Mfuna Endam and Dr Cormier contributed equally to this work. Ms Mfuna Endam and Dr Desrosiers had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Mfuna Endam, Cormier, Bossé, Filali-Mouhim, and Desrosiers. Acquisition of data: Mfuna Endam, Cormier, and Desrosiers. Analysis and interpretation of data: Mfuna Endam, Cormier, Bossé, Filali-Mouhim, and Desrosiers. Drafting of the manuscript: Mfuna Endam and Desrosiers. Critical revision of the manuscript for important intellectual content: Cormier, Bossé, Filali-Mouhim, and Desrosiers. Statistical analysis: Filali-Mouhim. Obtained funding: Desrosiers. Administrative, technical, and material support: Mfuna Endam, Cormier, and Desrosiers. Study supervision: Mfuna Endam, Bossé, and Desrosiers.

Financial Disclosure: None reported.

Funding/Support: This study was supported by the Fondation Antoine Turmel (Dr Desrosiers). Dr Bossé is a research scholar from the Heart and Stroke Foundation of Canada.

Previous Presentation: This study was presented at the 2009 Annual Meeting of the American Academy of Allergy, Asthma, and Immunology; March 15, 2009; Washington, DC.

Additional Contributions: We thank the research physicians, students, and assistants for sample and data collections. McGill University, Université de Montréal, and Genome Quebec Innovation Centre provided assistance and expertise throughout the conception and development of the entire genetics of CRS effort.