Published ahead of print on April 22, 2004, doi:10.1165/rcmb.2003-0128OC
© 2004 American Thoracic Society DOI: 10.1165/rcmb.2003-0128OC Signal Transducer and Activator of Transcription 6 Haplotypes and Asthma in the Indian PopulationInstitute of Genomics and Integrative Biology, Delhi; Vallabhbhai Patel Chest Institute, Delhi University, Delhi; and All India Institute of Medical Sciences, Delhi, India Address correspondence to: Balaram Ghosh, Ph.D., Molecular Immunogenetics Laboratory, Institute of Genomics and Integrative Biology, Mall Road, New Delhi-110007, India. E-mail: bghosh{at}igib.res.in, bghosh_igib{at}yahoo.com
In this paper, we report for the first time the results of an investigation on the association of signal transducer and activator of transcription 6 (STAT6) with asthma in the Indian population. A novel polymorphic CA-repeat in the proximal promoter region [R1] and a previously identified CA-repeat in the 5'-untranslated region [R3] were genotyped, and haplotypes [R1_R3] were generated using PHASE software. The 16 repeat allele at the R1 locus was positively associated (P = 0.01) with asthma. The 15 and 16 repeat alleles at the R3 locus were positively (P < 10–4) and negatively (P < 10–5) associated with asthma, respectively. Further, the 17_15 (P = 0.0031) and 16_15 (P = 0.001) haplotypes were found to be positively associated with asthma, whereas 17_14, 24_16, and 23_16 were negatively associated (P < 10–5). It appears that the R3 and R1 loci together play a bigger role in asthma than either of them alone, and the R3 locus has a larger effect than the R1 locus. Although alleles at the R1 locus appeared to be associated with total serum immunoglobulin E level, the genotypes showed no association, and the R3 locus showed no effect. As no exonic variants of STAT6 are known as yet, repeat polymorphisms in the regulatory regions and their haplotypes could be important in deciphering the genetic role of STAT6 in asthma and atopy.
Abbreviations: confidence interval, CI • interleukin, IL • likelihood ratio, LR • odds ratio, OR • signal transducer and activator of transcription 6, STAT6 • T helper type 2, Th2 • untranslated region, UTR
Asthma is a chronic airway disease, affecting 15–18% of the world's population (1). It has been estimated that atopic asthma accounts for up to 11–12% of the total population, with lost workdays due to asthma and airway disorders totaling 34% in the Indian population (2). Atopic asthma is a T helper type 2 (Th2)-mediated disorder with cytokines, implicated in the deviation of the immune system toward atopicity. Increased levels of Th2 cytokines, such as interleukin (IL)-4, IL-5, and IL-13, lead to elevated total serum immunoglobulin (Ig) E levels, higher eosinophil recruitment, and bronchial hyperresponsiveness that ultimately culminate in asthma pathogenesis (3). Gene–gene and gene–environment interactions have been implicated in the development of asthma (4). Several genetic studies have shown multiple loci to be associated with the disease. Linkage studies, in various populations, have narrowed down the presence of susceptibility or disease genes to chromosomal locations such as 5q31–33, 11p13, 12q13–24, and 16p12.1 (5–9). Similar studies in various populations also suggest that ethnic differences exist in the set of susceptibility genes associated with asthma (10). Of these loci, 12q21–23 harbors the signal transducer and activator of transcription 6 (STAT6) gene (consisting of 23 exons spanning a region of 19 kbp), which is thought to be an important candidate gene (11). Case-control studies in the Japanese population have shown a dinucleotide repeat in the 5'–untranslated region (UTR) of this gene to be associated with asthma and atopic disorders (12, 13). However, this observation has not been confirmed in a more stringent study on a Caucasian sib-pair cohort (14).
STAT6 plays a major role in the initiation of signals from activated Th2 cells, specifically through IL-4 and IL-13 receptors (15–18). STAT6 has also been implicated in the differential expression of chemokines, such as eotaxin-1, eotaxin-2, and thymus and activation regulated chemokine (19, 20). It is expressed in activated T cells in response to anti-CD3 antibody, phorbol myristate acetate, and other mitogens (21). IL-4 Receptor Overall, the genetic and biochemical evidence suggests that STAT6 is a strong candidate gene for disease pathogenesis and/or susceptibility to atopic disorders including asthma. We had previously screened STAT6 for identifying functional polymorphisms in the sequences coding for N-terminus, C-terminus, and SH2 domains in the Indian population (27). We had identified a novel CA repeat region in the proximal promoter of STAT6 (denoted as R1). A previously identified CA repeat in the 5'-UTR was also shown to be polymorphic in our population (denoted as R3) (27). In this study, we performed a case-control analysis to assess the association of these polymorphisms and their haplotypes with asthma and related phenotypes such as total serum IgE in the North Indian population. Our results suggest that the 15 and 16 CA repeat alleles at the R3 locus and the 16 repeat allele at the R1 locus were significantly associated with asthma, with confidence intervals (CI) of odds ratios (OR) being above 1.0 at 99% and 95% levels, respectively, for the two loci. Furthermore, the genotypes 15/17 at the R3 locus and 16/23 at the R1 locus were also found to be significantly associated with asthma. Log total serum IgE levels were found to be significantly associated with alleles at the R1 locus but not the R3 locus. Additionally, haplotypes containing 17 CA repeats at the R1 locus and 15 CA repeats at the R3 locus, and 16 repeats at R1 and 15 repeats at R3 locus were also associated with asthma. On the other hand, the 16 repeat allele at the R3 locus and the haplotypes 17_14, 23_16, and 24_16 were negatively associated with asthma. These results suggest that the R1 and R3 loci together play a bigger role in asthma than either of them alone and the R3 locus has a greater effect than the R1 locus.
Subjects The probands and controls were examined and recruited based on the evaluation of clinical, family, and environmental details using a standardized questionnaire as described previously (28). Patients (mean age 22.3 ± 12 yr) were diagnosed for asthma based on the National Asthma Education and Prevention Program (Expert Panel Report-II) guidelines and were examined for a self-reported history of breathlessness and wheezing. Asthma phenotype was established by pulmonary function test (FEV1, bronchial reversibility test  15% increase in FEV1 or FVC) using ß2-agonist inhaler (29). Probands were recruited with a positive family history of asthma and atopic disorders (positive skin prick test for common environmental allergens) with at least one first-degree relative afflicted with atopy-related disorders (28). Unrelated control subjects (mean age 25.2 ± 8 yr) were selected on the basis of no reported symptoms or history of allergic diseases or smoking. Those having self-reported histories of smoking and parasitic infestation were excluded from the study. The ethics committees of the participating centers and hospitals approved the study. Written informed consent was obtained from all individuals for performing a skin prick test and drawing blood samples. During interviews, the geographic region of origin and migration status of the patients and unrelated control subjects were also obtained. Total serum IgE was determined in the sera by sandwich enzyme-linked immunosorbent assay, as described previously (28).
Polymerase Chain Reaction Amplification of Polymorphic Regions
Statistical Analysis The Kolmogorov-Smirnov test was used to test for allelic association with disease at R1 and R3 loci (225 patients, 212 control subjects). Initial test for association between the R1 and R3 loci, stratified by phenotype (case, control), was done using a Cochran-Mantel-Haenszel test. Haplotypes were generated using the PHASE (http://archimedes.well.ox.ac.uk/pise/PHASE-simple.html, PHASE Ver. 2.0.2) program (30) for the patient (n = 225) and control (n = 212) groups. Default parameters were used to generate the haplotypes. No missing values were allowed. ORs were calculated and a Chi-square test for association with phenotype was performed for each of the 14 haplotypes having frequencies greater than 0.025, with Bonferroni correction (level of significance [  = 0.0035 to have overall = 0.05]). The repeats have been denoted with the allele size (16, 17, etc.), the genotypes with (16/17), and the haplotypes with (R1_R3 or 16_17). Equality of log total serum IgE levels in the patient and control groups was tested using the t test for equality of means. Association between the alleles (or genotypes) at the R1 and R3 loci with total serum IgE was tested for patients (n = 111) and control subjects (n = 126) using analysis of variance.
To determine the allelic frequencies of R1 and R3 loci in the STAT6 gene, genomic DNA from 225 patients and 212 unrelated control subjects were genotyped as described in MATERIALS AND METHODS. The allele ranges and frequencies obtained for the R1 and R3 loci have been summarized in Figures 2 and 3, respectively. The heterozygosity index for the R1 and R3 loci was found to be 0.85 and 0.75, respectively. The frequencies at the two loci were found to be consistent with Hardy-Weinberg equilibrium conditions in the control group. The mean log total serum IgE values for cases (n = 111, mean = 2.90 IU, SD = 0.75 IU) and control subjects (n = 126, mean = 2.54 IU,SD = 0.60 IU) are significantly different (t = 4.0, df = 235, P < 0.0001).
Association of R1 and R3 Loci with Asthma and Log (Total IgE Levels) The allele frequency distribution at the R1 locus appears to be bimodal. Interestingly, it is observed that there is an interchange of major and minor modes between the control subjects (major mode = 24 repeat and minor mode = 16 repeat) and patients (major mode = 16 repeats and minor mode = 24 repeats). The difference between patients and control subjects for the 16 allele was found to be significant (OR = 1.48, 95% [CI] = [1.09, 2.00] but 99% CI = [0.99, 2.20]; likelihood ratio [LR]  2 = 6.56, P = 0.01). However, the difference in the overall distribution of alleles in patients and control subjects was not found to be statistically significant (Kolmogorov-Smirnov 2 = 2.87, df = 2, P = 0.24) (Figure 2). One-way analysis of variance showed an association between the alleles at this locus and log total serum IgE levels in the patients (F = 2.65, df = [12, 221], P = 0.002). Tukey-Kramer Honestly Significant Difference showed significant difference between mean log IgE levels for these pairs of alleles: 16 and 25, 22 and 26, and 22 and 25. At the genotype level, 16/23 was found to be associated with phenotype with an OR of 2.20 and Wald's 95% CI (1.13, 4.25) (but 99% CI = [0.91, 5.23]). However, none of the genotypes at the R1 locus was associated with log total serum IgE levels (F = 1.20, df = [32, 110], P = 0.26).
For the R3 repeat locus, a significant difference in the allele count distribution was observed between control and patient groups (Kolmogorov-Smirnov
Association of STAT6 Haplotypes with Asthma
The haplotypes whose expected frequency was larger than 0.025 in either of the two groups are shown in Table 1. The odds in favor of patients rather than control subjects having 17_15 and 16_15 haplotypes were 2.63 with 99% CI = (1.08, 6.40) and 1.89 with 99% CI = (1.13, 3.13), respectively. The corresponding likelihood ratio
We had earlier reported a novel CA repeat polymorphism in the promoter region of the STAT6 gene (R1 locus) identified in the North Indian population (27). In this study, we found that this polymorphism is not associated with asthma. As seen from Figure 2, the number of allelic categories is large, but this is unlikely to affect the inference using the Kolmogorov-Smirnov test. However, the OR for the 16 repeat allele and the 16/23 genotype which are significant with 95% confidence but not 99% confidence, and the association of the R1_R3 haplotype with asthma, suggest that this locus deserves further study. Although we found a significant difference in mean log total serum IgE levels for different R1 alleles in patients, it does not enable us to obtain unique grouping of alleles. The other locus we examined in the STAT6 gene was R3. We found a significant association of the 15 CA repeat allele at the R3 locus with asthma but not with total serum IgE levels. In an earlier case-control study in the Japanese population, the R3 locus had been found to be associated with asthma (13 repeat allele) but not with total serum IgE levels, which is similar to our result (13). In contrast, in a sib-pair study in a German population, the R3 locus was not found to be associated with asthma but weak associations were observed for the total serum IgE levels and the eosinophil counts with the 17 and 16 repeat alleles, respectively (14). The difference between our result and that in the German study may be attributed to ethnic differences and to the differences in the study designs. For example, when we compared the allele frequencies at the R1 and R3 loci, and their haplotypes, in normal control samples obtained from North Indian (Asian Caucasian) and South Indian (Dravidian) populations, we observed that their distributions were significantly different (unpublished data). The results obtained from the Cochran-Mantel-Haenzel test on the genotypes of the R1 and R3 loci suggested that they were in linkage disequilibrium with each other. To confirm this, we needed to find the phases of the two loci on the chromosomes. As we could not distinguish phase from our random sample data, we used a Bayesian method for inferring haplotypes (30). The inferred 17_15 and 16_15 haplotypes were found to be significantly associated with asthma. Interestingly, it may be noted that the strength of the association of these haplotypes was stronger than for the R3 = 15 allele alone, as seen from the ORs (2.63 and 1.89 versus 1.76, respectively) or R1 alone (17 allele is not associated and 16 allele is associated at just over 1% level of significance). The negatively associated haplotypes 23_16, 24_16, and 17_14 also lead to the conclusion. This suggests that the combined effect of both the loci plays a bigger role, with the R3 locus being more important than the R1 locus. It would appear that the haplotype frequencies are very low and hence our results may be unreliable. In all, 91 different haplotypes are estimated by PHASE, resulting in small frequencies of individual haplotypes (expected average frequency of each haplotype being less than 0.011). We have tested for significance of only those haplotypes whose frequencies are more than 0.025 in at least one of the two groups; hence an independent study may not increase the haplotype frequencies and give more reliable results. Although there was a significant increase in the sample size (27), and PHASE was run on this data set independently, the frequencies of haplotypes generated from the control group in this study were found to be similar to those obtained previously. In all the haplotypes considered for analysis, the phase was declared uncertain by PHASE in eight control chromosomes (probability > 0.51) and in eleven patient chromosomes (probability > 0.65). In all other chromosomes analyzed (n = 225, Table 1), the probability of the inferred phase was greater than 0.90. Moreover, the presence of the 24_17, 23_17, and 17_15 haplotypes in two large extended families lends credibility to the haplotypes predicted using PHASE (Table 1). Nevertheless, the inference drawn from such haplotypic associations is only as good as the estimates of the predicted haplotypes. We have used a case-control study design, which has a higher chance than a family-based study, of finding a false-positive association. Because asthma is a complex disorder believed to be caused by the interaction of many genes, each having only a small effect, a case-control study may provide directions on promising loci. These leads need to be further tested using an independent cohort or well-controlled family-based study. However, due to small effects of individual genes, a family-based study may fail to detect significant association even when it exists (31). In this study, we have recruited only those patients who have at least one first-degree relative affected with atopy and/or asthma. As the case and control groups were age, sex, and ethnically and geographically matched, it is very unlikely that the association of the R3 locus with asthma could be a false positive. Indeed, no association was observed between the two groups for noncandidate chromosomal loci that were unlinked to asthma or atopic disorders (unpublished data). Therefore, it is very unlikely that our results are due to stratification or an inherent statistical bias. It is apparent that the use of the R1 and R3 polymorphisms in the generation of haplotypes in conjunction with single nucleotide polymorphism (SNP) data for this gene may yield more informative haplotypes. The haplotypes of SNPs obtained in the German population suggest that there may be a recombination hot spot in the gene but this putative region is located downstream of both the R1 and the R3 loci (14). Estimation of decay of linkage disequilibrium across the putative recombination hot spot would be important in deciphering functional aspects of this genomic region. In any event, if functional polymorphisms are present on the chromosomal background of specific R1_R3 haplotypes, then haplotypes that describe parts of the STAT6 gene flanking the putative recombination hot spot may provide a better association with asthma and total IgE. However, this hypothesis remains to be tested. As shown by other groups and ours, it is important to note that no coding variants of STAT6 gene were found (9, 14, 27). Thus, it appears that it is the alteration in the expression of STAT6 rather than its protein sequence that plays a regulatory role in asthma. In this context, both the R1 and R3 polymorphisms seem to be biologically relevant. Using promoter deletion analysis it has been shown that these loci are flanked by the critical transcription factor binding sites TFIIIA and the TATA box (32). Dinucleotide repeats are known to bind various minor groove-binding proteins, which can interact with the basal transcriptional complex to modulate transcription. It has been shown that dinucleotide repeats have a propensity for forming Z-DNA–like structures and that in the promoter regions these are capable of regulating transcription, for example, in the rat nucleolin gene (33). Similarly, the 5'-UTR is known to regulate translation of various genes through interaction with protein factors or by pseudoknot formation (34, 35). Therefore, it may be hypothesized that both the R1 and the R3 loci may modulate the expression of STAT6. This suggests that STAT6 could be an important modifier locus for regulating the atopic phenotypes. In conclusion, we report the association of the R3 locus and the novel R1_R3 haplotypes with asthma in the Indian population. Similar studies on the haplotypic (R1_R3) association in diverse populations could be beneficial in understanding the genetic contribution of STAT6 in the pathogenesis of asthma and atopic disorders.
The authors gratefully acknowledge the help provided by Drs. Jai Prakash Rishi, Mitali Mukherjee, Harish K. Pemde, Vipul Shandilya, Rajesh Chetiwal, Ms. Mamta Sharma, Shilpy Sharma, and Sangeeta Sharma. They thank the Council for Scientific and Industrial Research and Department of Biotechnology, Government of India, for financial assistance. Mr. Rana Nagarkatti acknowledges the fellowship granted by University Grants Commission, Government of India. The authors also acknowledge the Functional Genomics Unit of I.G.I.B, Delhi, for help in sequencing and genotyping. They are indebted to Prof. Carole Ober, University of Chicago, for her advice and critical review of the manuscript. Received in original form April 9, 2003 Received in final form March 17, 2004
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Top
Abstract
Introduction
immunoglobulin gene and causes the expression of the 
, CAAT enhancer binding protein delta; CCAAT Enhancer, cis-acting element required for the recruitment of transcription factors and in the assembly of the transcription complex.
