Introduction

Complex genetic diseases are highly prevalent in the general population and have a very significant socioeconomic impact on the health-care system. Asthma is the most common chronic childhood disease in developed nations.1 In India, the prevalence of asthma in adults is 2.38%2 and in children it varies from 6 to 31%.3, 4 Overall, it affects nearly 155 million people worldwide.5 Asthma is defined as a chronic inflammatory disorder of the airways, in which many cells and cellular elements have a role. The symptoms, resulting from airway hyperresponsiveness, include recurrent episodes of wheezing, breathlessness, chest tightness and coughing—especially at night or in the early morning.6 These symptoms are often associated with airflow obstruction within the lungs and are reversible—either spontaneously or by therapy.6

ADAM33 was first identified as a susceptibility gene for asthma by positional cloning in the year 2002.7 ADAM33 is located on chromosome 20p13, and its expression is restricted largely to mesenchymal cells, including fibroblasts and smooth muscle cells. The ADAM family of proteins, characterized by the presence of a disintegrin and metalloprotease domain (ADAM), have a role in cell–cell and cell–matrix interactions,8 as well as in cell migration,9, 10 cell–cell adhesion11 and signal transduction.12 The ADAMs protein family constitutes a variety of cell surface proteins, including growth factors, cytokines and receptors.8 ADAM33 is predicted to have protease activity and a domain structure composed of a signal sequence, prodomain, metalloprotease domain (with a zinc-binding motif), disintegrin domain, cysteine-rich domain (with an EGF-like motif), a transmembrane domain and a cytoplasmic domain.13 A number of reports have appeared since the first study documenting the positive association of ADAM33 polymorphism with asthma in the Caucasian population. Associations between distinct single-nucleotide polymorphisms (SNPs) of ADAM33 and various subphenotypes of asthma as well as bronchial hyperresponsiveness have been reported.7, 14, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34 Several SNPs in ADAM33 have been shown to be significantly associated with asthma, suggesting an important role for ADAM33 in the etiology of asthma.7, 14, 19, 20, 32 However, most data available for ADAM33 and asthma have been obtained from Caucasians, but little data on ADAM33 SNPs associated with asthma and atopic disease are available for Asian populations.14, 19, 21, 26, 33, 34 In light of this, we attempted to examine the association of ADAM33 gene polymorphisms F+1 G/A, S2 G/C, ST+4 A/C, ST+5 C/T and V4 C/G with asthma in Indian children.

Materials and methods

Subjects

The study group consisted of 211 cases and 137 controls. They were recruited between August 2007 and September 2009 from the Department of Pediatrics, Chattrapati Shahuji Maharaj Medical University, Lucknow, Uttar Pradesh, India. This study was approved by the institutional ethics committee and written informed consent for participation was obtained from parents/guardians of all the patients. On screening, the patient was suspected to have asthma if one or both of the following criteria were positive: (a) current diagnosis of asthma by the treating pediatrician and (b) the patient was prescribed antiasthma medication in the past. The diagnosis of suspected asthma and the severity of an asthma attack were then confirmed according to the Global Initiative for Asthma criteria (GINA) guidelines.6 Asthmatic patients were grouped into the following categories: (i) intermittent, (ii) mild persistent, (ii) moderate persistent and (iii) severe persistent. We excluded patients who had a history of other atopy-related diseases such as allergic rhinitis and diagnosed pulmonary disease tuberculosis. The inclusion criteria for controls were as follows: (1) no symptoms or history of asthma or other pulmonary diseases; (2) no symptoms or history of atopy; and (3) absence of first-degree relatives with a history of asthma or atopy.

Methods

DNA extraction and genotyping

Genomic DNA was extracted from 3 ml of frozen whole blood using a DNA Extraction Kit (FlixiGene DNA Kit, QIAGEN GmbH, Hilden, Germany) according to the manufacturer's protocol and following the manufacturer's instructions. We selected five SNPs in the ADAM33 gene (Figure 1), which were shown previously to have an allelic and/or haplotype association with asthma, including F+1(rs511898), V4(rs2787094), ST+4(rs44707), S2 (rs528557) and ST+5(rs597980). Using the unique rs accession numbers, SNP details and sequence data are available through NCBI databases http://www.ncbi.nlm.nih.gov. Genotyping was carried out by polymerase chain reaction restriction fragment length polymorphisms (PCR-RFLP) (Supplementary figure 1). Using isolated genomic DNA as template, PCRs were carried out. Each PCR was performed in a total reaction volume of 15 μl, with 20 pmol of each primer sequence. PCR products were digested at 37 °C overnight with 1 U of restriction enzymes. Detailed information on PCR conditions, primer sequences, restriction enzymes and so on is summarized in Table 1 . Nearly 10% of the samples were randomly selected and genotyped once again to rule out any inconsistency.

Figure 1
figure 1

Schematic representation of the ADAM33 gene on chromosome 20. (a) Chromosome 20 showing ADAM33 gene position 20p13. (b) Region covered by five genotyped polymorphisms (SNPs) in Mbs (megabases) and Kbs (kilobases). (c) Position of the genotyped polymorphisms (SNPs) in the ADAM33 gene with respect to the 22 exons (black) and untranslated regions (gray) of the gene.

Full size image
Table 1 Description of the investigated ADAM33 SNPs
Full size table

Statistical analysis

EpiINfo6 (available from the centers for Disease control and Prevention (Atlanta, GA, USA; http://www.cdc.gov/epo/epi/epiinfo.htm)) and SPSS 11.5 (Chicago, IL, USA) were used for statistical analysis. The χ2-test was used to determine differences in genotype/allele frequencies and deviation from Hardy–Weinberg equilibrium. Logistic regression was used to calculate odds ratio (OR). Homozygous genotype for the normal allele of each SNP in the control group was used as reference in calculating OR and 95% confidence intervals.16 Continuous variables were expressed as mean±s.d. The significance level for all statistical tests was set at P<0.05. P-values were corrected (Pcorr) for multiple corrections (Bonferroni's correction) in case of further subgrouping or stratification. The 0.05/N threshold was set up according to the Bonferroni's correction to account for multiple testing issues. N is the number of tested markers (genotypes) for each gene polymorphism. Therefore, for applying Bonferroni's correction, we have multiplied the P-value by the number of comparisons (for example, in case of a polymorphism having three genotypes, we multiplied the P-value by 3 and for carrier analysis the P-value was multiplied by 2). Case–control analysis was carried out to detect possible gene–environment interaction. Fisher's exact test was carried out to avoid type 1 error in the subgroup analyses due to the smaller sample size.16 Statistical analyses for the haplotype estimation and linkage disequilibrium (LD) were carried out using the expectation–maximization algorithm (Arlequin software version 2.00).15 The SNPAnalyzer (version 1.0; ISTECH; Istech, Kyungkido, Republic of Korea) was used to calculate D’ value, the measure index of LD.16

Results

This study was conducted from August 2007 to September 2009. Included were 211 cases of asthma, of which 51 (24.2%) were intermittent, 127 (60.2%) were mild persistent and 33 (15.6%) were moderate persistent. No cases of severe persistent asthma were detected. A total of 137 controls were included in the study. Table 2 shows the selected characteristics of the study population. Compared with controls, cases were likely to have father and grandfather or other family members with a present smoking habit. Proximity to an industrial area and road near a highway were the factors likely to cause a risk of developing disease asthma (P-value <0.05). Details regarding environmental risk factors for cases and controls were recorded and analysis was performed for further possible gene–environment interactions.

Table 2 Distribution of selected characteristics of children
Full size table

Association of ADAM33 gene polymorphisms with asthma compared with controls, cases with positive family history and total hospitalization

The genotype and allele frequencies of the ADAM33 gene polymorphisms in patients and healthy controls are shown in Table 3 . All five SNPs genotyped were in Hardy–Weinberg equilibrium (P>0.05) in both cases and controls. After comparing genotype and allele frequencies between patients and controls, increased risk for asthma was observed with the heterozygous genotype (GC), homozygous mutant genotype (CC) and mutant allele (C) of S2 SNP, with the heterozygous genotype (CT), homozygous mutant genotype (TT) and mutant allele of ST+5 SNP (T), the homozygous mutant genotype (CC) and mutant allele (C) of ST+4 SNP, the homozygous mutant genotype (GG) and mutant allele (G) of V4 SNP and with the homozygous mutant genotype (AA) and mutant allele (A) of F+1 SNP. F+1, V4 and ST+4 heterozygous genotypes (GA, CG and AC, respectively) were not significantly distributed among cases and controls.

Table 3 Association of five polymorphisms in ADAM33 with asthma in Indian children
Full size table

A case-only analysis was performed to observe the association between family history of asthma and total hospitalization rates. Among all five ADAM33 gene polymorphisms studied (F+1, V4, ST+4, S2 and ST+5), the homozygous mutant genotype of ST+5 SNP, the heterozygous genotype of ST+4, as well as V4, showed a high total hospitalization rate (Supplementary data file 1; P<0.05). The homozygous mutant genotype of ST+5 was observed to be significantly different among cases with a positive family history (OR=1.77 95%, confidence interval: 0.98–3.19), P=0.04), whereas the heterozygous genotype (AC) of ST+4 SNP observed a marginally significant difference among cases with a positive family history (OR=1.72 95% confidence interval: 0.95–3.11, P=0.0531; Supplementary data file 2).

Association of ADAM33 gene polymorphisms with severity of asthma

The genotype and allele frequencies of ADAM33 gene polymorphisms in subgroups of asthma patients and in healthy controls are shown in Table 4 . After comparing genotype and allele frequencies between asthma severity subgroups and controls, individuals with the homozygous mutant genotype (AA) of F+1 SNP, the homozygous mutant genotype (CC) of ST+4 SNP and the heterozygous and homozygous mutant genotypes (GC and CC) of S2 SNP were at increased risk for mild intermittent asthma. Polymorphisms V4 and ST+5 were not significantly associated with mild intermittent asthma. Individuals with the homozygous mutant genotype (AA) of F+1 SNP, the homozygous mutant genotype (GG) of V4 SNP, the heterozygous and homozygous mutant genotypes (GC and CC) of S2 SNP and the heterozygous and homozygous mutant genotypes (CT and TT) of ST+5 SNP were at increased risk for mild persistent asthma. Genotypes of ST+4 SNP were not statistically significantly associated with mild persistent asthma. Individuals with the heterozygous genotype (GC) of S2 SNP and the homozygous mutant genotypes of ST+5 and ST+4 SNPs (TT, CC) were at risk for moderate persistent asthma. SNPs F+1 and V4 were not significantly associated with moderate persistent asthma. Significant associations were observed with subgroups of asthma severity, but no single SNP was consistently associated with increasing severity of asthma across all subgroups.

Table 4 Association of ADAM33 gene polymorphisms with severity of asthma
Full size table

Haplotype analysis of ADAM33 polymorphisms

For statistical advantage, haplotypes with a frequency of <1% were excluded, as shown in Table 5 . In asthma patients, the frequencies of haplotype combinations AGCCT, GGACT and AGCCC were higher in patients and may be responsible for the risk of developing disease asthma, whereas frequencies of haplotype combinations ACAGT, AGCGC, AGCGT, GCAGC and GCCGT were lower in patients, which shows protective association with asthma. For all cases and controls, LD-measured D’ was high between most of the tested polymorphisms in the ADAM33 gene. (Supplementary data file 3) .A total of 10 pairs were observed, out of which 7 pairs, namely, F+1/ST+4, F+1/ST+5, V4/S2, V4/ST+5, ST+4/S2, ST+4/ST+5 and S2/ST+5, were in strong LD; in controls, 6 pairs, namely, F+1/V4, F+1/ST+5, V4/ST+4, V4/S2, V4/ST+5 and S2/ST+5, were in strong LD (Yates's corrected P value <0.05 and ∣D′∣≠0).

Table 5 Distribution of ADAM33 gene SNPs haplotypes and association with risk of Asthma
Full size table

Interaction of genotype with environmental risk factors

A case and control study was separately conducted to reveal any possible gene–environment interactions. Environmental risk factors included (1) type of road categorized into three subgroups—road of occasional traffic, busy road with vehicular traffic and residential area near highway; and (2) presence of industrial factor near residential area. Although all factors were prominent in patients as compared with controls, none of the genotype–environment risk factor interactions significantly modulated risk for asthma. Only ST+4 genotypes were significantly distributed among type-of-road subgroups. Near-highway-predicted percentages for the homozygous normal genotype (AA), the heterozygous genotype (AC) and the homozygous mutant genotype (CC) of ST+4 SNP were 11.8, 29.4 and 58.8%, respectively. Busy road with vehicular traffic-predicted percentages for homozygous normal genotype (AA), heterozygous genotype (AC) and homozygous mutant genotype (CC) of ST+4 SNP were 0.0, 40.0 and 60.0%, respectively. Road of occasional traffic-predicted percentages for homozygous normal genotype (AA), heterozygous genotype (AC) and the homozygous mutant genotype (CC) of ST+4 SNP were 22.0, 47.0 and 31.1%, respectively. For all three types of road subgroups compared with ST+4 genotypes, Fisher's exact P-value was 0.002, which indicates that an area of heavy traffic is a risk factor that initiates gene–environment interaction with asthma. (Insignificant data not shown regarding gene–environment interactions.)

Discussion

The present case–control study was conducted to assess the association of specific ADAM33 gene polymorphisms (F+1(rs5118980), V4(rs2787094), ST+4(rs44707), S2(rs528557) and ST+5(rs597980)) with childhood or childhood-onset asthma in children aged 1–15 years and in age- and sex-matched healthy controls. During the analysis of cases and controls, we found that S2 and ST+5 SNPs were significantly associated with asthma, whereas F+1, ST+4 and V4 homozygous mutant genotypes and mutant alleles were significantly different compared with controls.

Several groups have examined the association of ADAM33 gene polymorphisms (F+1, V4, ST+4, S2 and ST+5) with asthma in various populations and the findings are summarized in Table 6 . Similar to our results, S2 gene polymorphism was reported to be associated with asthma in a UK-only population,7 in a family-based study on Germans,19 as well as in African-American, US White and US Hispanic,20 Japanese21 and Thai32 populations. The ST+5 gene polymorphism was associated with a family-based study conducted by Werner et al.19

Table 6 Association of five genotyped ADAM33 gene polymorphisms in different populations
Full size table

The ST+4 ADAM33 gene polymorphism was found to be associated with US/UK combined and UK-only populations,7 in a German family-based study conducted by Werner et al.,19 as well as in Thai case–control32 and Japanese family-based studies.33 We also found a significant association of the ST+4 homozygous mutant genotype and mutant allele with asthma.

In our study, we found a significant association of the homozygous mutant genotype and mutant allele of F+1 SNP with asthma. Similar to our result, the F+1 gene polymorphism has shown positive association in a UK-only population and in German family-based studies.7, 19 Recently, Blakey et al.29 conducted transmission disequilibrium and case–control studies, followed by a meta-analysis of all existing data. The meta-analysis showed that F+1 and ST+7 (not performed in this study), which were not associated with asthma in most of the studies, were significantly associated with asthma.

The V4 gene polymorphisms of the ADAM33 gene have shown association in US/UK combined, UK-only, Dutch White and Chinese Han populations.7, 14, 20 We also found significant association of the ST+4 homozygous mutant genotype and mutant allele with asthma in our Indian population.

However, in contrast to our study, very large studies conducted by Lind23 and Raby et al.24 investigating childhood asthma could not find any association between single ADAM33 SNPs or haplotypes and childhood asthma. No association was found between asthma and ADAM33 in Korean, Australian, Chinese, German, Japanese and Caribbean Coast of Colombian populations.21, 25, 27, 28, 29, 31 A recent report by Bijanzadeh and colleagues also failed to find an association between asthma and the T1 SNP of the ADAM33 gene in a South Indian population.34 This may be because of allelic heterogeneity, a single disorder caused by different mutations within a gene. Another explanation is the difference in LD patterns between populations, which means that unobserved causal variants are the same in different populations, but that SNPs in strong LD with variants differ between populations or that the disease-causing variants are in strong LD with an SNP in one ethnic population, but not in others. Differences in the frequency of SNPs in cases and controls in different studies were summarized by Postma et al.30

Comparison of our results with those of various studies referred to in Table 6 indicates that no single SNP was universally associated with the asthma phenotype. One explanation for these differences might be the heterogeneity of asthma. There may be genetic differences among subtypes of asthma, such as pediatric, allergic/non-allergic asthma and so on. Asthma was not defined by common criteria in these investigations.

Haplotypes are considered to carry information about possibly unobserved causal variants in the region33 in asthma patients; the frequencies of haplotype combinations AGCCT, GGACT and AGCCC were higher in patients, suggesting that these three haplotypes increase the risk of asthma. Furthermore, we found that frequencies of haplotypes ACAGT, AGCGC, AGCGT, GGAGC and GCCGT were lower in patients compared with healthy controls. This result may imply a protective role of these five haplotypes in combination with asthma in Indian children. Although ADAM33 might have a key role in asthma through airway wall remodeling, its exact role in disease susceptibility is unclear. In previous studies, polymorphisms in mRNA were shown to contribute to transcript stability.35, 36 Therefore, it is likely that the susceptible allele present on the coding region of the ADAM33 transcript might affect its stability, which in turn might modulate the inflammation process. Our data strongly support the importance of ADAM33 in asthma across ethnic boundaries, and further investigations of the connections between genotypes and the functional roles of ADAM33 would be helpful to clarify the etiology of asthma. Given the recent advances in SNP genotyping, it is certain that more SNPs of this gene will be identified.

Our data suggest evidence for an association of the ADAM33 polymorphism with asthma risk. In conclusion, our study observed the significant association of SNPs S2 G/C and ST+5 C/T of the ADAM33 gene heterozygous and mutant genotypes and allele frequencies with asthma risk and significant association of F+1 G/A, ST+4 A/C, V4 C/G mutant genotypes and alleles with increased asthma risk. Frequencies of haplotype combinations AGCCT, GGACT and AGCCC were higher in patients and may cause risk of developing asthma, whereas frequencies of haplotype combinations ACAGT, AGCGC, AGCGT, GCAGC and GCCGT were lower in patients, which shows protective association with asthma. Additional studies with a larger sample size and with long-term follow-ups will undoubtedly lead to a more thorough understanding of the role of the ADAM33 gene polymorphism in childhood asthma in the Indian population.