SEARCH

SEARCH BY CITATION

Keywords:

  • asthma;
  • children;
  • Chinese;
  • genome-wide association;
  • single-nucleotide polymorphism

Abstract

  1. Top of page
  2. Abstract
  3. Patients and methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Conflicts of interest
  8. References
  9. Supporting Information

Background:  Single-nucleotide polymorphism (SNP)-based genome-wide association study revealed that markers on chromosome 17q21 were linked to childhood asthma but not atopy in Caucasians, with the strongest signal being detected for the SNP rs7216389 in the ORMDL3 gene. Such association was unknown in Chinese. This study delineated the allele and genotype frequencies of 10 SNPs at chromosome 17q21, and investigated the relationship between these SNPs and asthma and plasma IgE in southern Chinese children.

Methods:  Asthmatic children and non-allergic controls were recruited from pediatric clinics. Their plasma total and aeroallergen-specific IgE concentrations were measured by immunoassay. Ten SNPs on 17q21 region were genotyped by multiplex SNaPshot™, and their genotype associations with asthma traits analyzed using multivariate regression.

Results:  315 patients and 192 controls were enrolled. The allele frequency for C allele of rs7216389 varied significantly from 0.232 in our controls, 0.389 in Han Chinese to 0.536 in Caucasians. Asthma diagnosis was associated with rs11650680 and five other SNPs including rs7216389 (= 0.019–0.034), whereas atopy was associated only with rs11650680 (= 0.0004). Linear regression revealed the covariates for plasma total IgE to be significant for rs11650680 (= 0.008–0.0002). Haplotypic associations were found with atopy and increased plasma total IgE, with the respective odds ratios and 95% confidence intervals for TTTCCGTT haplotype to be 0.21 and 0.09–0.52 (= 0.0002) and 0.41 and 0.18–0.90 (= 0.025).

Conclusion:  Childhood asthma and atopy are associated with chromosome 17q21 in Chinese, but such association may involve genes other than ORMDL3 in this region.

Asthma is one of the most common childhood diseases in developed countries. About one-tenth of Hong Kong schoolchildren suffer from asthma (1). Most patients have enhanced production of immunoglobulin E (IgE) to a variety of allergens. Asthma is a complex disease involving interactions between many genetic and environmental factors (2, 3). Over the past decade, genetic linkage and association studies have identified more than 25 asthma or allergy susceptibility loci. Nonetheless, fewer than half of these associations have been successfully replicated in multiple populations. Recently, Moffatt et al. published the first genome-wide association (GWA) study for childhood asthma in a cohort of 994 European patients and 1243 controls (4). They identified highly significant associations between asthma and single-nucleotide polymorphisms (SNPs) on chromosome 17q21, and replicated these results in two independent Caucasian cohorts. In particular, the strongest signal was detected for SNP rs7216389 in the gene ORM1-like Protein 3 (ORMDL3). Following additional experiments, they identified ORMDL3 to be a novel candidate gene for asthma.

Subsequently, several groups tried to replicate these genetic associations. In view of the great interethnic variations of asthma burden in the United States, Galanter et al. studied the association between asthma and ORMDL3 in three independent cohorts of Mexicans, Puerto Ricans, and African-Americans (5). They observed significant associations between two SNPs within ORMDL3 (rs4378650 and rs12603332) and asthma in Mexicans and African-Americans. None of the eight tested SNPs was associated with baseline lung function or bronchodilator responsiveness. Another Japanese study of the SNP rs7216389 supported the importance of ORMDL3 for asthma (6). Very recently, Tavendale et al. reported that the SNP rs7216389 in ORMDL3 was also associated with the risks for asthma susceptibility and exacerbations among Scottish children and young adults (7). These consistent findings suggest that ORMDL3 or adjacent genes on chromosome 17q21 is a strong risk factor for asthma in ethnically diverse populations. Nonetheless, inconsistent results at the SNP level call for more replication studies to define the genetic associations with markers at this chromosomal region.

We hypothesized that asthma and plasma IgE are associated with polymorphic markers at chromosome 17q21 region in Chinese, but that these traits may be linked with ORMDL3 or its adjacent gene(s). The objectives of this study were to delineate the genetic epidemiology of SNPs at chromosome 17q21 in Chinese children, and to investigate the associations between asthma traits and these SNPs in our population.

Patients and methods

  1. Top of page
  2. Abstract
  3. Patients and methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Conflicts of interest
  8. References
  9. Supporting Information

Study population

Unrelated Chinese children with physician-diagnosed asthma and aged 5 to 18 years were enrolled from pediatric clinics of a university teaching hospital. The diagnosis of asthma was made according to the American Thoracic Society (ATS) guidelines (8). Non-allergic controls were recruited among children attending our hospital for minor non-respiratory complaints, and all of them did not report any allergic disease using the validated International Study of Asthma and Allergies in Childhood (ISAAC) questionnaire (1). All cases and controls were free of any infectious symptom for 4 weeks before study. The Clinical Research Ethics Committee of our University approved these studies, and subjects and their parents gave written informed consent.

Plasma IgE measurements

Total IgE concentration in plasma was measured by micro-particle immunoassay (IMx analyser; Abbott Laboratories, Abbott Park, IL, USA) and specific IgE antibodies to Dermatophagoides pteronyssinus, cat and cockroaches were measured by fluorescent enzyme immunoassay (AutoCAP; Phadia AB, Uppsala, Sweden). Data on total IgE concentrations were compared with our local references (upper limits of normal: 120 kIU/l for 5 to 6 years; 160 kIU/l for 6 to 7 years; 180 kIU/l for ≥ 8 years) (9). Specific IgE concentration ≥ 0.35 kIU/l was considered positive, and subjects with one or more allergen-specific IgE were considered atopic.

Spirometric measurements

All asthmatics underwent spirometry (COMPACT II; Vitalograph, Buckingham, UK) to measure their forced expiratory volume in 1 s (FEV1) and forced vital capacity (FVC). Results were compared with local references (10).

Selection and genotyping of SNPs at chromosome 17q21

The published SNPs at chromosome 17q21 were reviewed from HapMap data (11), and seven tagging SNPs were selected from those with minor allele frequency (MAF) ≥0.05 for Han Chinese and according to r 0.90. The three SNPs with the most significant P-values from Moffatt et al. (4) were added to our panel, i.e. total 10 SNPs, that is the maximum number of SNPs that can be genotyped by the SNaPshot assay. Table 1 summarizes the locations and gene functions of these SNPs, and Fig. 1 describes their distribution with respect to different genes at the 17q21 region.

Table 1.   Details of our 10 single-nucleotide polymorphisms on chromosome 17q21
SNPGene Role (12)rs number (11)Probe length (bp)
  1. SNP, single-nucleotide polymorphism; UTR, untranslated region.

T+34578CZNFN1A3Intron 2381647019
T+10507C/T+3620CORMDL3/GSDMLDownstream/Intron 1721638923
T+3033C/T-3854CORMDL3/GSDML3′ UTR/Promoter227889426
T-2357C/T-9262CORMDL3/GSDMLIntron 1/Promoter1760892540
C-3894TORMDL3Promoter374424656
C-4929AORMDL3Promoter479540235
650352538
G-10587AGSDM1Promoter100765442
C+6707T/C-8576TGSDM1/PSMD3Intron 6/Promoter385919246
C-5193TTOP2APromoter1165068050
image

Figure 1.  The distribution of our SNPs along the chromosome 17q21 region.

Download figure to PowerPoint

Genomic DNA was extracted from peripheral blood leukocytes. The genomic region of interest at chromosome 17q21 was amplified by multiplex polymerase chain reactions (PCRs), the details of which are summarized in Table S1. These PCR reactions were performed in a total volume of 18 μl containing 120 ng DNA, 200 μM dNTPs, 1X PCR Gold buffer with 2.5 mM MgCl2, 5–10 pmol of each primer and 0.5 unit of AmpliTaq Gold DNA polymerase (Roche Molecular Systems, Branchburg, NJ, USA). Samples were denatured at 95°C for 10 min followed by 30 cycles of 95°C for 60 s, 58°C for 60 s, followed by 72°C for 75 s, and a final extension at 72°C for 10 min.

Our 10 SNPs were genotyped by SNaPshot Multiplex kit (Applied Biosystems, Foster City, CA, USA) as previously described (13). In brief, 5 μl of mixed PCR products from the above reactions were then incubated with 2 μl of exonuclease I/shrimp alkaline phosphatase (ExoSAP-IT; USB, Cleveland, OH, USA) at 37°C overnight, followed by heating at 80°C for 15 min to inactivate ExoSAP-IT. Following this, 2.5 μl of SNaPshot Multiplex kit, 1.5 μl of ExoSAP-IT treated PCR products and 1 μl of mixed probes (Table S2) were incubated for 25 cycles at 96°C for 10 s, 54°C for 5 s, and 60°C for 30 s. Finally, 0.6 unit of SAP (USB) was added to the reaction product and incubated at 37°C for an hour to remove the unused primers and unwanted PCR products. SAP was then denatured by heating at 75°C for 15 min. The final reaction mix containing 9.25 μl of Hi-Di Formamide, 0.5 μl of sample and 0.25 μl of GeneScan-120 LIZ internal size standard (Applied Biosystems) was denatured at 95°C 5 min, and genotypes of our 10 SNPs were identified by different fluorescent signals by ABI-3130 Genetic Analyzer (Applied Biosystems). Finally, the multiplex SNaPshot results from 40 randomly selected subjects were confirmed by direct sequencing of the target chromosome 17q21 region using BigDye Terminator Cycle sequencing kit (Applied Biosystems).

Statistical analysis

Results were expressed as proportions or mean and standard deviation (SD). Patient characteristics were compared using the Student t-test or anova for numerical data, and chi-squared or Fisher’s exact test for categorical variables. Plasma total IgE was logarithm-transformed to base 10 before analysis. Hardy–Weinberg equilibrium (HWE) for the 10 SNPs in cases and controls was evaluated by the exact test. To investigate the pattern of linkage disequilibrium (LD) in our SNPs on chromosome 17q21, pairwise LD coefficient was calculated for each SNP pair by Haploview (Daly Lab, Cambridge, MA, USA). The associations between our SNPs and dichotomous phenotypes were analyzed using multivariate logistic regression, and with quantitative traits using linear regression. Two alternative models (recessive and co-dominant) were fitted for each outcome, adjusting for age and gender as covariates. Haplotypes were assigned as maximum likelihood estimates using EH program, and their associations with asthma traits were analyzed after 1000 permutations using WHAP (14). All comparisons were made two-tailed, and P-values less than 0.05 were considered significant.

Results

  1. Top of page
  2. Abstract
  3. Patients and methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Conflicts of interest
  8. References
  9. Supporting Information

Study population

Three hundred and fifteen patients and 192 controls were recruited. Table S3 summarizes their demographic and clinical characteristics. The mean (SD) age of patients and controls was 11.1 (3.9) years and 11.8 (4.1) years, respectively (= 0.074). Patients had higher log-transformed plasma total IgE than controls [mean (SD): 2.60 (0.61) vs 1.86 (0.70); < 0.0001]. Atopy was also more common among our cases (86.6%) than the controls (43.2%; < 0.0001).

Characterization of SNPs at chromosome 17q21

The overall genotyping efficiency was 99.6%. Table 2 summarizes MAF for our 10 SNPs. Two SNPs did not follow HWE: rs2278894 among the cases (= 0.0002) and rs4795402 among the controls (= 0.019), and they were excluded from genotype and haplotype analyses. The LD scores for these SNPs among our subjects were generally medium to high for SNPs around the ORMDL3 gene (Fig. S1), but only one haplotype block with rs17608925, rs3744246 and rs4795402 could be assigned to our SNP panel. Two SNPs distant from ORMDL3 (rs3859192 in GSDM1 and rs11650680 in TOP2) were in generally low LD with the other SNPs.

Table 2.   Interethnic comparisons of minor allele frequencies for polymorphic markers at chromosome 17q21 in our subjects with HapMap and published data for non-allergic controls
SNPMinor allelePresent studyHapMap data (12)Published data (4–7)
AsthmaticsControlsHan ChineseJapaneseEuropeanAfricanCaucasianJapaneseAfrican-American
  1. SNP, single-nucleotide polymorphism.

rs3816470C0.1960.2210.3890.2730.5250.583
rs7216389C0.1970.2320.3890.2890.5170.1250.503–0.5360.2870.216
rs2278894C0.0250.0210.089000
rs4795402A0.1860.2270.3560.2670.3170.483
rs6503525C0.6290.5700.4330.5110.4750.517
rs17608925C0.0560.0730.0890.0560.1580.033
rs1007654A0.2400.2840.4330.3440.3080.241
rs3859192T0.5430.5080.4220.6110.5000.3500.409–0.4640.321
rs11650680T0.2060.2580.2000.1330.23300.199–0.2120.002
rs3744246T0.2010.2300.3440.2670.2670.2920.197–0.2070.307

Asthma traits and individual polymorphisms

The relation between dichotomous asthma-related traits and SNPs in chromosome 17q21 was analyzed by logistic regression (Tables 3 and 4). Adjusted for age and gender as covariates, asthma diagnosis was significantly associated with rs3816470, rs7216389, rs11650680 and rs3744246 under the co-dominant model, and with rs6503525 and rs11650680 under the recessive model. Atopy was significantly associated only with rs11650680 under both co-dominant (= 0.0004) and recessive (= 0.0004) models. Plasma total IgE level was analyzed as a continuous variable. The covariate for this variable was significant for rs11650680 under both co-dominant and recessive models (= 0.008 and 0.0002 respectively), suggesting that subjects homozygous for the T allele had lower plasma total IgE levels (Table 5). The SNP rs3859192 was again marginally associated with plasma total IgE.

Table 3.   Association between physician-diagnosed asthma and SNP genotypes in our subjects
GenotypeAsthmatics (n = 315)Controls (n = 192)Recessive model*Co-dominant model*
n (%)n (%)OR (95% CI)POR (95% CI)P
  1. CI, confidence interval; OR, odds ratio; NA, not available; Ref, reference.

  2. *Analyzed by logistic regression, adjusted for age and gender as covariates.

rs3816470
 TT203 (65.1)110 (57.3)1.00Ref1.00Ref
 TC96 (30.8)79 (41.1)0.67 (0.44–0.95)0.025
 CC13 (4.2)3 (1.6)3.04 (0.84–10.91)0.0892.60 (0.72–9.42)0.146
rs7216389
 TT203 (64.4)107 (55.7)1.00Ref1.00Ref
 TC100 (31.7)81 (42.2)0.64 (0.44–0.94)0.023
 CC12 (3.8)4 (2.1)2.05 (0.64–6.52)0.2251.74 (0.54–5.58)0.354
rs6503525
 GG50 (15.9)33 (17.2)1.00Ref1.00Ref
 GC134 (42.5)99 (51.6)0.86 (0.51–1.44)0.567
 CC131 (41.6)60 (31.3)1.55 (1.06–2.27)0.0241.39 (0.81–2.38)0.234
rs17608925
 TT280 (89.5)163 (85.3)1.000Ref1.00Ref
 TC31 (9.9)28 (14.7)0.62 (0.36–1.08)0.092
 CC2 (0.6)0 (0)NA0.999NA0.999
rs1007654
 GG182 (58.0)96 (50.5)1.00Ref1.00Ref
 GA113 (36.0)80 (42.1)0.76 (0.52–1.11)0.160
 AA19 (6.1)14 (7.4)0.87 (0.42–1.78)0.6960.77 (0.37–1.62)0.492
rs3859192
 CC69 (22.0)43 (22.6)1.000Ref1.00Ref
 CT149 (47.5)101 (53.2)0.92 (0.58–1.45)0.712
 TT96 (30.6)46 (24.2)1.32 (0.87–1.99)0.1911.24 (0.74–2.09)0.418
rs11650680
 CC196 (62.2)108 (56.8)1.00Ref1.00Ref
 CT108 (34.3)66 (34.7)0.92 (0.62–1.36)0.676
 TT11 (3.5)16 (8.4)0.42 (0.19–0.94)0.0340.41 (0.18–0.92)0.031
rs3744246
 CC202 (64.3)106 (56.1)1.00Ref1.00Ref
 CT98 (31.2)79 (41.8)0.63 (0.43–0.93)0.019
 TT14 (4.5)4 (2.1)2.36 (0.76–7.36)0.1392.00 (0.63–6.29)0.238
Table 4.   Association between atopy † and SNP genotypes in our subjects
GenotypeAtopy (= 340)Non-atopic controls (= 148)Recessive model*Co-dominant model*
n (%)n (%)OR (95% CI)POR (95% CI)P
  1. CI, confidence interval; OR, odds ratio; Ref, reference.

  2. *Analyzed by logistic regression, adjusted for age and gender as covariates.

  3. †Defined as ≥ one allergen-specific IgE in plasma.

rs3816470
 TT213 (63.2)87 (58.8)1.00Ref1.00Ref
 TC112 (33.2)58 (39.2)0.80 (0.53–1.21)0.298
 CC12 (3.6)3 (2.0)1.53 (0.42–5.59)0.5191.41 (0.38–5.20)0.604
rs7216389
 TT215 (63.2)84 (56.8)1.00Ref1.00Ref
 TC114 (33.5)60 (40.5)0.74 (0.49–1.12)0.153
 CC11 (3.2)4 (2.7)1.10 (0.34–3.56)0.8700.98 (0.30–3.21)0.979
rs6503525
 GG52 (15.3)30 (20.3)1.00Ref1.00Ref
 GC152 (44.7)73 (49.3)1.24 (0.72–2.12)0.439
 CC136 (40.0)45 (30.4)1.50 (0.99–2.28)0.0591.75 (0.99–3.10)0.055
rs17608925
 TT298 (88.2)127 (86.4)1.00ref1.00ref
 TC39 (11.5)19 (12.9)0.85 (0.47–1.54)0.588
 CC1 (0.3)1 (0.7)0.48 (0.03–7.80)0.6080.47 (0.03–7.66)0.598
rs1007654
 GG194 (57.2)73 (50.0)1.00Ref1.00Ref
 GA123 (36.3)63 (43.2)0.75 (0.49–1.13)0.165
 AA22 (6.5)10 (6.8)0.92 (0.42–2.01)0.8330.81 (0.36–1.81)0.607
rs3859192
 CC69 (20.4)39 (26.7)1.00Ref1.00Ref
 CT174 (51.3)69 (47.3)1.45 (0.89–2.37)0.136
 TT96 (28.3)38 (26.0)1.12 (0.71–1.75)0.6341.44 (0.83–2.50)0.201
rs11650680
 CC213 (62.6)82 (56.2)1.00Ref1.00Ref
 CT118 (34.7)48 (32.9)0.95 (0.62–1.46)0.820
 TT9 (2.6)16 (11.0)0.21 (0.09–0.50)0.00040.20 (0.09–0.49)0.0004
rs3744246
 CC216 (63.1)82 (56.6)1.00Ref1.00Ref
 CT113 (33.3)58 (40.0)0.74 (0.49–1.13)0.161
 TT12 (3.5)5 (3.4)0.95 (0.33–2.77)0.9220.85 (0.29–2.51)0.765
Table 5.   Association between SNP genotypes and plasma total IgE as a quantitative trait
GenotypeAll subjects (= 498)Plasma log10IgE*P-values
n (%)anovaRM†CDM†
  1. anova, analysis of variance; CDM, co-dominant model; RM, recessive model.

  2. *Expressed in mean ± standard deviation.

  3. †Analyzed by linear regression, adjusted for age and gender as covariates.

rs3816470
 TT308 (62.2)2.36 ± 0.720.1300.3900.325
 TC172 (34.7)2.23 ± 0.77
 CC15 (3.0)2.52 ± 0.71
rs7216389
 TT305 (61.2)2.35 ± 0.720.3770.7710.313
 TC178 (35.7)2.26 ± 0.77
 CC15 (3.0)2.39 ± 0.73
rs6503525
 GG82 (1.16)2.20 ± 0.800.2370.0990.166
 GC228 (45.8)2.33 ± 0.72
 CC188 (37.8)2.36 ± 0.72
rs17608925
 TT435 (87.9)2.33 ± 0.740.7380.6090.630
 TC58 (11.7)2.26 ± 0.71
 CC2 (0.4)2.55 ± 1.08
rs1007654
 GG273 (55.2)2.34 ± 0.730.4840.2650.311
 GA190 (38.4)2.31 ± 0.74
 AA32 (6.5)2.18 ± 0.80
rs3859192
 CC111 (22.4)2.22 ± 0.800.0430.0270.017
 CT246 (49.7)2.30 ± 0.73
 TT138 (27.9)2.44 ± 0.69
rs11650680
 CC301 (60.7)2.36 ± 0.720.0010.00020.008
 CT170 (34.3)2.32 ± 0.76
 TT25 (5.0)1.79 ± 0.63
rs3744246
 CC303 (61.3)2.36 ± 0.720.2130.5040.300
 CT174 (35.2)2.25 ± 0.77
 TT17 (3.4)2.45 ± 0.73

Asthma traits and haplotypes at chromosome 17q21

Eight-locus haplotypes were assigned from SNPs that followed HWE and with MAF ≥5%. There were no significant association of haplotypes with asthma diagnosis (= 0.453). On the other hand, atopy was significantly associated with our eight-locus haplotypes (= 0.048) There was also a marginally significant haplotypic association (= 0.053) for increased plasma total IgE levels. Table 6 summarizes the post hoc analyses for individual haplotypes. The haplotype TTTCCGTT was significantly protective against both atopy and increased plasma total IgE, with the respective OR and 95% CI for these phenotypes being 0.21 and 0.09–0.52 (= 0.0002) and 0.41 and 0.18–0.90 (= 0.025).

Table 6.   Relationship between atopy and increased plasma total IgE levels and haplotypes* assigned from eight SNPs at chromosome 17q21 region
PhenotypeHaplotypeFrequencyOR95% CIP
AtopyNo atopy
  1. CI, confidence interval; OR, odds ratio; Ref, reference haplotype used for comparison.

  2. *Haplotypes were constructed from the following SNPs: rs3816470, rs7216389, rs17608925, rs3744246, rs6503525, rs1007654, rs3859192 and rs11650680. Only haplotypes with ≥ 0.05 frequencies in at least one group were listed.

  3. †Yates-corrected P-values obtained by chi-squared or Fisher exact test.

  4. ‡Compared with the upper limits of local references (120 kIU/l for 5–6 years; 160 kIU/l for 6–7 years; 180 kIU/l for ≥8 years).

Atopic statusTTTCCGTC0.3880.2521.00Ref
TTTCCGCC0.1140.0910.780.36–1.690.620
TTTCCGTT0.0380.1160.210.09–0.520.0002
CCTTGACC0.0480.0820.370.15–0.930.032
TTTCCGCT0.0420.0660.390.15–1.040.063
TTTCGGCC0.0410.0670.390.15–1.040.063
TTTCGACC0.0500.0520.600.22–1.640.388
  IncreasedNormal   
Plasma total IgE‡TTTCCGTC0.3690.2771.00Ref
TTTCCGCC0.1240.0970.960.49–1.920.958
TTTCCGTT0.0530.0960.410.18–0.900.025
CCTTGACC0.0480.0780.480.20–1.130.102
TTTCCGCT0.0400.0630.480.19–1.240.147
TTTCGGCC0.0360.0670.410.16–1.050.065
TTTCGACC0.0510.0470.860.33–2.260.908
TTTCGACT0.0280.0500.390.13–1.140.091

Discussion

  1. Top of page
  2. Abstract
  3. Patients and methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Conflicts of interest
  8. References
  9. Supporting Information

Using case–control test of genetic associations, this study found significant associations between asthma and atopy traits and a number of SNPs on chromosome 17q21 in southern Chinese children. Interestingly, the SNP rs7216389 most significantly associated with asthma in the Caucasian and Japanese populations was only marginally associated with asthma in this study. On the other hand, we observe that asthma, atopy and plasma total IgE were consistently associated with rs11650680 in TOP2A located upstream of ORMDL3. In this replication study, the associations between polymorphic markers at chromosome 17q21 and asthma-related traits are consistently significant or close to significant at 5% alpha level. These findings are clearly not a case of ‘chance significant finding’ coming out of a bunch of statistical comparisons (15), indicating a substantial evidence for the effects of chromosome 17q21 markers on asthma and atopy. We consider that adjusting for multiple comparisons is not needed. Besides, the majority of genetic associations at least for rs11650680 would remain significant after such corrections. On haplotypic analysis, atopy and increased plasma total IgE also showed significant associations with an eight-locus haplotype in this chromosomal region. These findings strongly support the importance of genes in chromosome 17q21 in conferring the risks of childhood asthma and atopy.

The initial publication of asthma GWA by Moffatt (4) identified the chromosome 17q21 as a candidate locus for childhood asthma in Caucasians. They went on to show that this genetic association was due to the functional SNP rs7216389 and others in ORMDL3 on this chromosome region. Two of the three replication studies published to date focused mainly on rs7216389 (6, 7), which could not provide data on the SNPs of other genes on chromosome 17q21. The third study genotyped eight ORMDL3 SNPs in three independent cohorts of Mexicans, Puerto Ricans, and African-Americans (5). They observed significant associations between two SNPs within ORMDL3 (rs4378650 and rs12603332 but not rs7216389) and asthma in Mexicans and African-Americans. The present study also detected weak yet significant associations between asthma and two SNPs (rs7216389 and rs3744246) within ORMDL3. These consistent findings suggest that ORMDL3, or adjacent genes on chromosome 17q21, is a strong risk factor for asthma in ethnically diverse populations. Despite the fact that ORMDL3 is expressed ubiquitously, its function or pathogenic role in human diseases remains unclear. Hjelmqvist et al. showed ORMDL3 to have significant sequence conservation to homologues in Drosophila and Saccharomyces cerevisiae (16). The Japanese group demonstrated that ORMDL3 expression in normal human lung fibroblasts was strongly induced by polyinosine-polycytidylic acid, one of the pathogen-associated microbial patterns, but not by lipopolysaccharide, lipopeptide 2 or Pam3CSK4. This finding suggests that ORMDL3 may be important in defense against viral respiratory infections (6).

We observe from the above findings that asthma is associated with different SNPs within ORMDL3 in different ethnic groups – with rs7216389 in Japanese, Scottish and our southern Chinese and with rs4378650 and rs12603332 in Mexicans and African-Americans. This discrepancy of the results at SNP level suggests that gene(s) adjacent to ORMDL3 on chromosome 17q21 may also influence asthma susceptibility. For instance, our study as well as the asthma GWA (4) identified a strong association signal for rs11650680 of TOP2A that is about 500 kbp from rs7216389. As these two SNPs fell within a large LD block in the Caucasian population, Galanter et al. (5) argued that such genetic association implied that these SNPs were in LD with one or more of the causative SNPs. However, our LD pattern (Fig. 1) showed that rs11650680 and rs7216389 are not linked in Chinese in contrast to what Moffatt et al. found in their Caucasian samples. rs11650680 is located in the promoter region of TOP2A (DNA topoisomerase II, alpha isozyme; OMIM# 126430) (12). This gene is mapped to chromosome 17q21.2, and encodes a DNA topoisomerase of 29 327 base pairs with 35 exons (17). Topoisomerase II from eukaryotic cells catalyzes the relaxation of supercoiled DNA molecules, catenation, decatenation, knotting, and unknotting of circular DNA, implying its importance in cell cycle and DNA repair. TOP2A is increasingly being recognized to be an important predictor of treatment responses for many cancers (18, 19). Nonetheless, there has not been any study on the relevance of TOP2A in asthma and atopy pathogenesis. Together with our consistent findings of rs11650680 with a number of asthma traits, we suggest that TOP2A may be an additional candidate gene for asthma and plasma IgE. Further genetic studies with larger sample size and dense SNP mapping of the locus around rs11650680 are needed to confirm our observation. As in our published findings between IL13, IL4RA and TARC (20), it would also be interesting to investigate the epigenetic interactions for asthma and atopy between genes on chromosome 17q21 and other known candidate genes.

This study provides further data to support the interethnic variations of MAF for polymorphic markers on chromosome 17q21 (Table 2). The minor C allele of rs7216389 in our southern Chinese (0.232) was similar to that of Japanese (0.287) but much lower than Han Chinese (0.389) and Caucasians (0.503 to 0.536). In contrast, MAF for rs11650680 and its adjacent SNP (rs3859192) were similar to the published data for Caucasians (4, 11). Such interethnic differences possibly also affect the haplotype structure at chromosome 17q21: our initial HapMap search assigns rs3744246 and rs4795402 to the adjacent haplotype blocks for Caucasians, but these two SNPs are clustered into a single haplotype block in our southern Chinese (Fig. S1). These differences in MAF also affected the power of this study to detect significant genetic associations. For example, C allele of rs7216389 is the ‘wild-type’ allele in Caucasians but becomes a minor allele in our Chinese subjects. This finding is similar to what we found for CD14, DEFB1, and PTGDR in Chinese children (13, 21, 22). With MAF that is 30% lower than that found by Moffatt (4), this study in Chinese would expectedly yield less impressive P-values for rs7216389 with asthma traits. Barnes reviewed that the genetic epidemiology of asthma candidate loci showed marked variations among different racial and ethnic groups (23). As the global burden of asthma and allergy also varied widely (24), it would be crucial to delineate the influence of chromosome 17q21 markers for asthma traits in different populations.

One limitation of this study relates to the use of different criteria for defining cases and controls. Asthma in our cases was diagnosed by the standard ATS criteria (8), i.e. presence of suggestive asthma symptoms together with BHR or reversibility of airflow limitation. On the other hand, our controls were ascertained by the Chinese translated ISAAC questionnaire (1). They gave negative returns for all four items ‘current wheeze’, ‘asthma ever’, ‘rhinoconjunctivitis ever’ and ‘eczema ever’. de Meer et al. found that 66 of 470 asymptomatic schoolchildren had BHR to hypertonic saline according to the ISAAC protocol (25). Without measuring lung functions for the non-allergic controls, we agree that some of them may have asymptomatic BHR. This limitation would thus reduce the differences in genotype frequencies of our studied SNPs between cases and controls for the outcomes. Nonetheless, we argue that the findings of genetic associations with our SNPs would suggest that they are more likely to be genuine. Another limitation is that the number of controls was much lower than that of our cases, and this issue may weaken the power of this study in detecting any significant genetic association. Despite this, the existing sample size has 80% power to detect with 95% confidence an odds ratio of 1.8 for asthma for any SNP with MAF 0.2, which was applicable for eight of our studied SNPs (i.e. MAF for controls in Table 2). In addition, our controls were recruited from non-allergic children who attended our hospital outpatient clinics for minor, non-respiratory complaints. Although we cannot exclude any selection bias or population stratification, the pattern of allergen sensitizations of these controls (see Table S3) are comparable with those of local schoolchildren (26). Despite this, this study may be improved by recruiting control subjects from our general population.

In conclusion, this case–control study supports the genetic associations between asthma and related traits and chromosome 17q21 region. However, these clinical phenotypes were most strongly linked to rs11650680 of the TOP2A gene that is distant from rs7216389, the SNP in ORMDL3 that is published among the Caucasian population. Future studies should replicate our findings and also investigate the possible interactions between genes on chromosome 17q21 in determining the susceptibility for asthma and atopy.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Patients and methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Conflicts of interest
  8. References
  9. Supporting Information

We thank E. Yung and Y. S. Wong for performing spirometry and processing DNA samples in our subjects. This study was supported by the Research Committee Group Research Scheme (no. 3110034) and Direct Grant for Research, The Chinese University of Hong Kong.

Conflicts of interest

  1. Top of page
  2. Abstract
  3. Patients and methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Conflicts of interest
  8. References
  9. Supporting Information

The sponsors of the study had no involvement in the study design; in the collection, analysis, and interpretation of data; in the writing of the report; and in the decision to submit the paper for publication.

References

  1. Top of page
  2. Abstract
  3. Patients and methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Conflicts of interest
  8. References
  9. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Patients and methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. Conflicts of interest
  8. References
  9. Supporting Information

Figure S1. Linkage disequilibrium (LD) pattern among our 10 SNPs by Haploview analysis.

Table S1. Primers for the three multiplex PCR reactions.

Table S2. Mixed probes for multiplex genotyping using SNaPshot kit.

Table S3. Clinical and laboratory characteristics of our subject.

Please note: Wiley-Blackwell are not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.

FilenameFormatSizeDescription
ALL_1873_sm_FigS1.ppt51KSupporting info item
ALL_1873_sm_TablesS1-S3 and FigS1.doc261KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.