Potential association of DCBLD2 polymorphisms with fall rates of FEV(1) by aspirin provocation in Korean asthmatics.

Aspirin exacerbated respiratory disease (AERD) is a clinical syndrome characterized by chronic rhinosinusitis with nasal polyposis and aspirin hypersensitivity. The aspirin-induced bronchospasm is mediated by mast cell and eosinophilic inflammation. Recently, it has been reported that the expression of discoidin, CUB and LCCL domain-containing protein 2 (DCBLD2) is up-regulated in lung cancers and is regulated by transcription factor AP-2 alpha (TFAP2A), a component of activator protein-2 (AP-2) that is known to regulate IL-8 production in human lung fibroblasts and epithelial cells. To investigate the associations between AERD and DCBLD2 polymorphisms, 12 common variants were genotyped in 163 AERD subjects and 429 aspirin tolerant asthma (ATA) controls. Among these variants, seven SNPs (rs1371687, rs7615856, rs828621, rs828618, rs828616, rs1062196, and rs8833) and one haplotype (DCBLD2-ht1) show associations with susceptibility to AERD. In further analysis, this study reveals significant associations between the SNPs or haplotypes and the percentage of forced expiratory volume in one second (FEV(1)) decline following aspirin challenge using multiple linear regression analysis. Furthermore, a non-synonymous SNP rs16840208 (Asp723Asn) shows a strong association with FEV(1) decline in AERD patients. Although further studies for the non-synonymous Asp723Asn variation are needed, our findings suggest that DCBLD2 could be related to FEV(1)-related phenotypes in asthmatics.

INTRODUCTION

Aspirin-exacerbated respiratory disease (AERD), is a clinical syndrome with inflammatory responses in both the upper and lower respiratory tracts (1). It is characterized by chronic rhinosinusitis with nasal polyposis which is followed by asthma and hypersensitivity to aspirin (2). Aspirin-induced bronchospasm in asthma patients are mediated by mast cell and eosinophilic inflammation which is already ongoing before the first aspirin ingestion (2, 3). The mechanisms for pathogenesis of AERD are not completely explained, however, it has been suggested that over-production of cysteinyl leukotriene (CysLT) and its receptor on inflammatory cells occurs in the respiratory tract of AERD patients (4).

The transmembrane protein encoded by discoidin, CUB and LCCL domain-containing protein 2 (DCBLD2) gene, also abbreviated as CLCP1 or ESDN, is regarded as a contributor for regulation of cell proliferation. Our previous genome-wide and follow-up studies showed nominal associations between DCBLD2 polymorphisms and AERD (5). Despite little information on the function of DCBLD2, this gene has also been elucidated to be up-regulated in the development and metastasis of lung cancers (6). In addition, the transcription factor TFAP2A induces an inhibitory effect on DCBLD2 transcription (7). The TFAP2A, an important transcription factor, is a component of activator protein-2 (AP-2) that regulates IL-8 expression in human lung fibroblasts and epithelial cells (8). On the other hand, previous studies have also proposed that the up-regulation of DCBLD2 could be involved in pathways related to immune and injury-mediated remodeling (9, 10). In addition, neuropilin-1, whose domains are structurally similar with DCBLD2 domain proteins and has been identified as an isoform of a specific vascular endothelial growth factor receptor in human airway epithelial cells, showed a higher expression in patients who have chronic rhinosinusitis and/or nasal polyposis (9, 11).

Based on the possible relations of DCBLD2 to airway remodeling, we investigate further the associations of DCBLD2 single nucleotide polymorphisms (SNPs) with the fall rates of forced expiratory volume by aspirin provocation as well as AERD development under various genetic models.

MATERIALS AND METHODS

Study subjects

Asthmatic subjects were recruited from the Asthma Genome Research Center comprising hospitals of Soonchunhyang University in Seoul and Bucheon, Chungnam National University and Chungbuk National University in Korea. Patients met the definition of asthma based on the Global Initiative for Asthma guidelines 2010 (http://www.ginasthma.org/guidelines-gina-report-global-strategy-for-asthma.html). All subjects had a history of dyspnea and wheezing during the past year plus one of the following: 1) > 15% increase in forced expiratory volume in one second (FEV1) or > 12% increase plus 200 mL following inhalation of a short-acting bronchodilator, 2) < 10 mg/mL PC20 methacholine, or 3) > 20% increase in FEV1 following 2 weeks of treatment with inhaled steroids and long-acting bronchodilators. Twenty-four common inhalant allergens (e.g., dust mites, cat fur, dog fur, cockroaches, grasses, trees, ragweed pollen; Bencard Co. Ltd., Brentford, UK) were used in a skin-prick test. Total IgE was measured by the CAP system (Pharmacia Diagnostics, Uppsala, Sweden). Atopy was defined as a wheal reaction equal to or greater than histamine or 3 mm in diameter.

The oral aspirin provocation test was performed with slight modifications in increasing doses of aspirin (12), following the guideline of EAACI/GA2LEN (13). Briefly, the patient having history of aspirin hypersensitivity was given 30 mg and those having no history started 100 mg of aspirin orally. Symptoms, external signs (urticaria and angioedema), blood pressure and FEV1 were documented every 30 min for a period of 2 hr. In the absence of any symptoms or signs suggestive of adverse reaction after 2 hr, 60 mg or 100 mg of aspirin was administered and the same measurements were repeated every 1 hr, with doses of 450 mg until the patient developed a reaction. If no reaction occurred 5 hr after the final dose, the test was deemed negative. Changes in the FEV1 were followed for 5 hr after the final aspirin dose. Aspirin-induced bronchospasm, reflected by the rate (%) of decline in FEV1, was calculated as the pre challenge FEV1 minus the post challenge FEV1 divided by the pre-challenge FEV1. Subjects were categorized into two groups based on OAC reactions: ≥ 20% decrease in FEV1 or a 15%-19% decrease in FEV1 with naso-ocular reactions as AERD patients group, whereas a < 15% decrease in FEV1 without naso-ocular reactions as ATA controls group.

SNP selection and genotyping

We selected common SNPs based on the frequencies in Asian population from the International HapMap Project database (http://hapmap.ncbi.nlm.gov/index.html.en). The selected 12 SNPs were genotyped in a total of 592 asthmatic subjects composed of 163 AERD and 429 ATA subjects. Genotyping was carried out using TaqMan assay in the ABI prism 7900HT sequence detection system (Applied Biosystems, Carlsbad, California, USA) with the assessment of data quality by duplicate DNAs (n = 10). Genotype data were obtained using the ABI-PRISM sequence detection system (SDS) software version 2.3. SNPs that did not match the following standards were excluded from the study: 1) a minimum call rate of 95%; 2) no duplicate error; and 3) P values of Hardy-Weinberg Equilibrium more than 0.05. All the 12 SNPs of DCBLD2 were successfully genotyped.

Statistics

The association of SNPs and haplotypes of DCBLD2 with AERD was carried out with logistic analyses controlling for age, sex, smoking status, atopy and body mass index (BMI) as covariates using the Statistical Analysis System (SAS). The FEV1 change induced by aspirin provocation which was considered as a continuous variable was subjected to a simple linear regression analysis, and the differences in the values among the genotypes or haplotypes were examined using a linear regression model that controlled for age, sex, atopy and smoking status as covariates. For linkage disequilibrium (LD), we examined Lewontin's D' (|D'|) and the LD coefficient r2 between all pairs of biallelic loci using the Haploview v4.1 software downloaded from the Broad Institute (http://www.broadinstitute.org/mpg/haploview) (14). Haplotypes were first estimated using PHASE software (15), and then computed by logistic analyses using SAS.

Ethics statement

The institutional review board approved the protocol (SCHBC_IRB_05_02), and all subjects provided informed consent.

RESULTS

Characteristics of study subjects

The clinical profiles of the study subjects are summarized in Table 1. All of the subjects were asthma patients and were divided into two groups, AERD patients and ATA control groups, according to degree of aspirin sensitivity. Overall, a decrease in FEV1 of -15% to 68% induced by aspirin provocation was observed. The percentage of decrease in FEV1 by aspirin provocation in AERD patients (24.6%) were significantly higher compared to ATA controls (3.5%; P < 0.001). The values of predicted FEV1 %, PC20 methacholine and body mass index in AERD patients were significantly lower than those of ATA controls (P = 0.009, 0.02 and 0. 001, respectively). In addition, the mean age of a first medical examination in AERD was significantly lower than ATA (P = 0.001). Percentage of smoking status in AERD patients was also lower compared to ATA controls (P = 0.02).

Table 1

Clinical profiles of study subjects

Clinical profile of AERD was compared to ATA controls. AERD, aspirin-exacerbated respiratory disease; ATA, aspirin-tolerant asthma.

Genotyping and haplotypes of DCBLD2 polymorphisms

We selected 12 common polymorphisms of DCBLD2, six in exon regions and six in introns, based on the frequencies on Asian population from the International HapMap Project (Table 2, Fig. 1). The minor allele frequencies (MAFs) of these 12 SNPs in the Korean asthmatics (n = 592) are shown in Table 2. The genotype distributions of all loci are in Hardy-Weinberg equilibrium (Table 2). The linkage disequilibrium coefficients (|D'|) among the SNPs were calculated for all of the study subjects (Fig. 1). Complete LDs were observed between two SNPs; rs7615856 and rs828621 (r2 = 1). Twelve haplotypes were constructed and five of them with frequencies over 0.05 (DCBLD-ht1 to DCBLD-ht5) were included in the association analysis (Fig. 1).

Table 2

Single nucleotide polymorphisms and minor allele frequencies of human DCBLD2 in Korean subjects

*P values of deviation from Hardy-Weinberg Equilibrium (HWE) in Korean population. MAF, minor allele frequency.

Fig. 1

Gene maps and haplotypes of DCBLD2. (A) Polymorphisms of DCBLD2 investigated in this study. Coding exons are marked by shaded blocks; 5'- and 3'-untranslated region (UTR) by white blocks. Two SNPs (rs7615856 and rs828621) are in complete LD (r2 = 1). (B) Haplotypes of DCBLD2 in the Korean population. Only those with frequencies over 0.05 are analyzed for associations. (C) LD blocks and correlation coefficients among DCBLD2 polymorphisms.

Associations of DCBLD2 polymorphisms with FEV1 decline and AERD development

Initially, this study investigated the associations between genotypes or haplotypes and the percentage of FEV1 decline following aspirin challenge using multiple linear regression analysis (Table 3). Interestingly, nine SNPs (rs2439224, rs1371687, rs7615856, rs828621, rs828618, rs828616, rs16840208, rs17270986 and rs8833) were significantly associated with the percentage FEV1 decline induced by aspirin challenge in the asthmatics (P < 0.05, Table 3). Furthermore, among the significantly associated SNPs, a non-synonymous SNP rs16840208 (D723N or Asp723Asn) was found to induce a strong genetic effect (P = 0.02 under co-dominant; P = 0.009 in dominant model). In the haplotype analysis, DCBLD2-ht1 and DCBLD2-ht4 were also associated with the percentage FEV1 decline induced by aspirin challenge in the asthmatics (P = 0.009 for DCBLD2-ht4 under dominant model to P = 0.02 for DCBLD2-ht1 and DCBLD2-ht4 under co-dominant, Table 3).

Table 3

Association of SNPs and haplotypes of DCBLD2 with FEV1 decline by aspirin provocation in asthmatics

Genotype distribution of each SNP is presented as the number of subjects (percentage of FEV1 decline by aspirin provocation, mean ± SE). P values for linear regression analysis controlling age, sex, smoking status and atopy as covariates. C/C, Major homozygote; C/R, Heterozygote; R/R, Minor homozygote.

Furthermore, results from regression analysis comparing AERD and ATA groups revealed that nine SNPs (P = 0.001 for rs828618 under dominant model to P = 0.05 for rs2439224, rs828618, rs17270986 under co-dominant) and two haplotypes (P = 0.001 for DCBLD2-ht1 under dominant model to P = 0.05 for DCBLD2-ht1 under dominant; P = 0.03 for DCBLD2-ht2 under recessive) were significantly associated with the percentage FEV1 decline induced by aspirin challenge in AERD patients (Supplementary Table 1). In contrast, findings from regression analysis of ATA group showed no significant association between DCBLD2 SNPs and FEV1 decline induced by aspirin provocation, except for DCBLD2-ht1 haplotype (P = 0.02 in recessive model, Supplementary Table 1).

In further association analysis between DCBLD2 SNPs/haplotypes and the risk of AERD using multiple logistic models, two SNPs (rs828618 and rs828616; P = 0.05 and P = 0.04, respectively) and one haplotype (DCBLD2-ht1; P = 0.05) showed nominal signals with the risk of AERD (Table 4).

Table 4

Analyses of association of DCBLD2 polymorphisms with risk of AERD

Logistic analyses controlling for age, sex, smoking status, atopy and body mass index (BMI) as covariates are performed using the Statistical Analysis System (SAS). OR (95% CI) and P values of co-dominant model are also reported in the supporting information of our previous report (Kim et al. 2010b). MAF, minor allele frequency; OR, odds ratio; CI, confidence interval.

DISCUSSION

To our knowledge, this study is the first to investigate the association between DCBLD2 and AERD. In this study, data from logistic and linear regression analyses showed significant associations of DCBLD2 polymorphisms with FEV1 decline by aspirin provocation (P = 0.009 for rs16840208 under dominant model to P = 0.04 for rs7615856, rs828621 and rs8833 under co-dominant, Table 3) and nominal signals to AERD development (P = 0.04 for rs828616 to P = 0.05 for rs828618 under co-dominant model, Table 4). Furthermore, the strength of association with the fall rate of FEV1 by aspirin provocation was increased in the AERD subgroup (P = 0.001 for rs828618 under dominant model to P = 0.05 for rs2439224, rs828618, rs17270986 under co-dominant, Supplementary Table 1) compared to ATA (P > 0.05), suggesting that genetic variations of DCBLD2 may affect decline of FEV1 by aspirin provocation in AERD patients.

Recently, a significant up-regulation of DCBLD2, previously identified as CLCP1, has been elucidated in the metastasis of lung cancer cell line (6), suggesting that DCBLD2 potentially plays a role in lung-related functions. In addition, DCBLD2, also known as ESDN, is a key factor in the modulation of vascular smooth muscle cell (VSMC) growth and regulation of VSMC proliferation processes that are involved in vascular remodeling. DCBLD2 is up-regulated in remodeling arteries, and its down-regulation by RNA interference (RNAi) has been found to significantly enhance VSMC DNA synthesis and migration through induction of platelet-derived growth factor (PDGF), a prototypic growth factor for VSMCs (10, 16).

The transcription factor TFAP2A, a member of the AP-2 family, plays an important role in regulation of DCBLD2 transcription. It has been shown that TFAP2A binding to DCBLD2 promoter region induces a decrease in promoter activity measured by luciferase assay in this gene. Conversely, mutations in TFAP2A binding site in DCBLD2 promoter lead to the increase in transcription activity (7). AP-2 is a regulator for IL-8 production that is associated with the destructive pulmonary inflammation and the recruitment of neutrophils. Deletion of AP-2 binding site in the upstream region of IL-8 promoter significantly decreased IL-8 transcription activity (8), suggesting that DCBLD2 might be related to the production of interleukins in human pulmonary diseases through interaction with AP-2. This possibility is supported by the facts that discodin domain affects immune responses in the involvement of interleukins (17, 18).

Although the functional properties of DCBLD2 in tumor development and vascular remodeling have been reported, to date no previous study has proposed the functional or genetic effects of DCBLD2 on asthma and aspirin-induced hypersensitivity. However, neuropilin-1, which is structurally similar to DCBLD2 (19) has been found to be related to the suppressor function of CD4+CD25+ T cells in airway inflammation and hyper-responsiveness (20). In addition, neuropilin-1 is associated with the development of nasal polyposis, one of the symptoms of AERD. Distribution of neuropilin-1, as a co-receptor of vascular endothelial growth factor (VEGF), was shown to be 7-fold higher in nasal lavage from patients with polyposis than control subjects. Moreover, blockade of VEGF, which is suppressed by neuropilin-1, results in increased apoptosis and inhibition of autocrine epithelial VGEF production (11). Therefore, with the structural similarity to neuropilin-1, there is a possibility DCBLD2 may play a similar role in aspirin sensitivity to asthma.

The bronchoalveolar lavage (BAL) in asthma has been implicated in asparagine levels. The BAL fluid from asthmatic patients showed higher level of asparagine (Asn, N) than from normal controls (21). Other studies have reported that a polymorphism in the human neuropeptide S receptor (NPSR) which produces an amino acid substitution from Asn to isoleusine (Ile) (Asn107Ile) is associated with susceptibility to asthma. The Asn107Ile polymorphism of NPSR was found to be highly expressed in human asthmatic airway tissue (22, 23). In the present study, rs16840208 (Asp723Asn) is significantly associated with FEV1 decline by aspirin provocation (P = 0.02 under co-dominant and 0.009 under dominant model), suggesting that the amino acid change from aspartic acid (Asp) to asparagine of rs16840208 may be a genetic factor affecting the risk of pulmonary diseases including asthma.

Among the SNPs associated with FEV1 decline by aspirin provocation and/or AERD, 5 SNPs (rs2439224, rs1371687, rs7615856, rs828621, and rs828618) were located in intronic region of DCBLD2. Despite difficulty in assessing the functions of SNPs that are not positioned at the exon and promoter regions, previous studies have suggested that intronic SNPs may play important roles in gene transcription rate and abnormal splicing events such as exon skipping, activation of cryptic splice sites and production of alternatively spliced isoforms in human disease phenotypes (24, 25). Furthermore, it has also been reported that intronic SNPs in solute carrier family 6, member 7 (SLC6A7) (26), solute carrier family 6, member 12 (SLC6A12) (27), kinesin family number 3A (KIF3A) (28), calcium channel, voltage-dependent, gamma subunit 6 (CACNG6) (29) and fibrous sheath interacting protein 1 (FSIP1) (30) are associated with asthma and/or decline in FEV1 by aspirin challenge.

In summary, findings from this study provide evidences that genetic polymorphisms of DCBLD2 including a non-synonymous rs16840208 (Asp723Asn) might induce decline in FEV1 by aspirin ingestion in Korean asthma patients. Furthermore, considering that 5-lipoxygenase, whose pathway is one of the central mechanisms to produce CysLTs from arachidonic acid (31), plays a role in VSMC that is modulated by DCBLD2 (10, 32), these findings suggest that genetic variants of DCBLD2 could provide a new strategy for the control of aspirin intolerance.