Published ahead of print on September 22, 2005, doi:10.1164/rccm.200503-505OC
© 2006 American Thoracic Society doi: 10.1164/rccm.200503-505OC
T-Bet Polymorphisms Are Associated with Asthma and Airway HyperresponsivenessChanning Laboratory, Division of Pulmonary and Critical Care Medicine, Department of Medicine, Brigham and Women's Hospital; Division of Pulmonary and Critical Care Medicine, Beth Israel Deaconess Medical Center; Harvard Medical School; Harvard School of Public Health; Veterans Administration Medical Center; Harvard Partners Center of Genetics and Genomics; Boston; The Broad Institute of the Massachusetts Institute of Technology and Harvard University, Cambridge, Massachusetts; Departments of Internal Medicine, Pathology, and Immunology, Washington University School of Medicine, St. Louis, Missouri; and Division of Molecular Life Sciences and College of Pharmacy, Ewha Women's University, Seoul, Korea Correspondence and requests for reprints should be addressed to Benjamin Raby, M.D.C.M., M.P.H., Channing Laboratory, Brigham and Women's Hospital, Boston, MA 02115. E-mail: benjamin.raby{at}channing.harvard.edu
Rationale: T-bet (TBX21 or T-box 21) is a critical regulator of T-helper 1 lineage commitment and IFN-  production. Knockout mice lacking T-bet develop airway hyperresponsiveness (AHR) to methacholine, peribronchial eosinophilic and lymphocytic inflammation, and increased type III collagen deposition below the bronchial epithelium basement membrane, reminiscent of both acute and chronic asthma histopathology. Little is known regarding the role of genetic variation surrounding T-bet in the development of human AHR. Objectives: To assess the relationship between T-bet polymorphisms and asthma-related phenotypes using family-based association. Methods: Single nucleotide polymorphism discovery was performed by resequencing the T-bet genomic locus in 30 individuals (including 22 patients with asthma). Sixteen variants were genotyped in 580 nuclear families ascertained through offspring with asthma from the Childhood Asthma Management Program clinical trial. Haplotype patterns were determined from this genotype data. Family-based tests of association were performed with asthma, AHR, lung function, total serum immunoglobulin E, and blood eosinophil levels. Main Results: We identified 24 variants. Evidence of association was observed between c.–7947 and asthma in white families using both additive (p = 0.02) or dominant models (p = 0.006). c.–7947 and three other variants were also associated with AHR (log-methacholine PC20, p = 0.02–0.04). Haplotype analysis suggested that an AHR locus is in linkage disequilibrium with variants in the 3'UTR. Evidence of association of AHR with c.–7947, but not with other 3'UTR SNPs, was replicated in an independent cohort of adult males with AHR. Conclusions: These data suggest that T-bet variation contributes to airway responsiveness in asthma.
Key Words: immunoglobulin E • single nucleotide polymorphism • T-box • TBX21
T-bet (TBX21, Online Mendelian Inheritance of Man database no. 604895) is a nuclear transcription factor belonging to the T-box gene family of DNA-binding proteins, and is a potent inducer of IFN-
Given these data, it is possible that genetic variation at the T-bet locus confers susceptibility to asthma, AHR, and atopy. Two reports evaluating T-bet single nucleotide polymorphisms (SNPs) in asthma populations from Korea and Finland failed to demonstrate evidence of a relationship with either a diagnosis of asthma or total serum IgE levels; however, both studies reported they were statistically underpowered to detect genetic associations of modest effect (odds ratios,
Populations The Childhood Asthma Management Program (CAMP) is a multicentered North American clinical trial designed to investigate the long-term effects of inhaled antiinflammatory medications in children with mild to moderate asthma (10, 11). A total of 968 of the 1,041 children enrolled in CAMP and 1,518 of their parents contributed DNA samples. A total of 580 complete nuclear families are included in the analysis presented here. See online supplement for additional details for inclusion criteria and phenotyping protocols. The Normative Aging Study (NAS) is a longitudinal study of aging established by the Veterans Administration in 1961 (12). The initial cohort consisted of 2,280 men from the greater Boston area ages 21 to 80 yr at the time of entry into the study between 1961 and 1969. Since entry, volunteers have reported for periodic examinations that include spirometric tests and modified methacholine challenge test. Methacholine responders were identified by a fall in FEV1 of 20% or more from baseline during the challenge test. Nonresponders demonstrated a less than 20% fall in FEV1. DNA for 200 responders and 436 nonresponders was available for genotyping. See online supplement for additional details.
Human Subjects
Polymorphism Discovery and Genotyping
Statistical Analysis FBAT version 1.5.3 (19, 20) was used for family-based association analysis with asthma in CAMP. Quantitative trait analysis was performed with principal components analysis as implemented in PBAT (FBAT-PC) (21, 22). Evidence for haplotype association was assessed with the likelihood-ratio score test implemented in TRANSMIT (23) (asthma) and FBAT (quantitative traits) (24). See online supplement for additional details. NAS data were analyzed with SAS Genetics (SAS Institute, Cary, NC) and used generalized linear models for analysis of methacholine responsiveness (log-transformed dose–response slope). Models were adjusted for age, pack-years of cigarette smoking, and current smoking status. See online supplement for additional details.
SNP Discovery Resequencing of the T-bet locus identified 24 SNPs, including six variants in the 22 people with asthma and 18 variants in the people without asthma (see Figure E1 in the online supplement). Only one nonsynonymous variant was identified (Ile339Val), observed once in an individual without asthma. Virtually none of the SNPs identified reside in highly conserved regions: none were identified in the highly conserved T-box domain, and the only SNP identified in a conserved region of the transcript, Gly130Gly, was synonymous. One upstream variant, c.–7947A > G, did localize to a region of modest phylogenetic conservation, although the minor allele at this locus is present in rodent sequence.
SNP Genotyping, Linkage Disequilibrium, and Haplotype Block Structure in CAMP
SNP Association Analysis Presented in Table 3 are the results of the single-SNP, family-based analysis of association with the asthma phenotype and with AHR among whites. Only 1 of 16 SNPs tested demonstrated evidence of association with asthma: c.–7947G was undertransmitted to affected offspring. Under the additive model, this distortion has a p value, uncorrected for multiple comparisons, of 0.02. Stronger evidence for association was found at this locus with a dominant model (p = 0.006), although such modeling was performed post hoc. No SNP mapping to the transcript itself demonstrated significant association with asthma, although the nonsynonymous His33Gln polymorphism demonstrated the most skewed transmission distortion for all SNPs tested, with a transmitted:untransmitted ratio of approximately 3:2 for the Gln allele (p = 0.28). Similar patterns of transmission distortion were observed when analysis was restricted to those probands with at least one positive skin-prick test. No association with asthma was found in either the African-American or Hispanic families, although sample size in these two cohorts was insufficient to detect associations of modest effect (data not shown).
Repeated measures of postbronchodilator FEV1 (10–12 measurements/participant), methacholine PC20 (2–5 measurements), total serum IgE levels, and blood eosinophil levels (2 measurements each) were obtained over the 4.5 yr of observation during the CAMP trial. In whites, no evidence for association was found with FEV1, IgE, or eosinophils. However, several SNPs across the T-bet locus were associated with AHR in CAMP, including c.–7947, one variant localized to the T-bet transcript (c.2122), and two 3' genomic variants (g.21126 and g.56983). Of these associations, those with c.2122 and g.21126 met conservative criteria for significance (rank-order Bonferroni corrected p values of 0.04 and 0.02, respectively). Because of the small number of informative families in other ethnic groups, quantitative analysis was restricted to the white cohort.
Haplotype Association Analysis
Replication of T-bet Association with AHR We attempted to replicate evidence for an association between T-bet variants and AHR in a second population: 200 methacholine- hyperresponsive (responders) and 436 nonresponsive men participating in the NAS (12). Baseline characteristics are presented in Table 5. Methacholine responders were older, were more likely to have previously smoked, had lower baseline lung function, had higher total serum IgE levels and blood eosinophil levels, and were more likely to have a positive skin test to one of four common aeroallergens. Six variants were genotyped: g.–19698, c.–7947, His33Gln, IVS1+7240, c.2122, and c.2314. All were in Hardy-Weinberg equilibrium. None of these SNPs differentiated responders from nonresponders (data not shown). Among the responsive subjects, a strong association was found between methacholine responsiveness and c.–7947 (Figure 2). The other five variants were not associated with AHR in NAS, although trends were observed for g.–19698 (p = 0.06 for common vs. rare homozygote genotypes). No associations with total serum IgE or blood eosinophil levels were observed in this cohort.
T-bet serves as a regulator of Th1 development both by inducing IFN- production and by inhibiting Th2 cytokines interleukin (IL)-4 and IL-5 (reviewed by Murphy and Glimcher [25]). The T-bet locus maps to chromosome 17q21, a region occasionally linked with both asthma and atopy-related phenotypes (26, 27). We provide data demonstrating a potential association between T-bet variants and AHR. One variant, c.–7947, was found to be weakly associated with asthma susceptibility, and four variants dispersed across the T-bet locus (c.–7947, c.2122, c.21126, and c.56983) were significantly associated with AHR in children with asthma. We have also provided some evidence that the relationship between c.–7947 and airway responsiveness is reproducible in a second cohort of older men with AHR, although confirmation of this and other associations in additional populations is warranted before firm conclusions regarding the generalizability and true impact of these findings can be drawn. No associations were found with FEV1, serum IgE levels, or blood eosinophil levels in either cohort. This study complements earlier work in which we demonstrated a significant pharmacogenetic effect of the His33Gln variant in the long-term treatment of asthma. The glutamine variant was associated with a marked improvement in AHR after treatment with inhaled corticosteroids (28). Interestingly, in the current study, this nonsynonymous variant demonstrated the greatest degree of transmission distortion to asthmatic offspring, with a transmitted:untransmitted ratio of approximately 3:2 for the glutamine allele. Because of the relatively small number of informative families (n = 54), this distortion did not reach statistical significance, and so it is not possible to claim this SNP as an asthma-susceptibility variant. Substantially larger cohorts would be needed to test this hypothesis. This study is the fourth to characterize the genetic variation of the T-bet locus and to test this variation in asthmatic cohorts. Chung and colleagues tested seven SNPs and one major haplotype in a Korean hospital-based cohort of 550 patients with asthma and 170 control subjects and found no evidence for association with asthma, serum IgE, or skin-test reactivity (6). Ylikoski and colleagues evaluated 15 SNPs in 120 asthma families from a founder population in central eastern Finland, and similarly did not detect evidence of association with either asthma or serum IgE levels (7). Akahoshi and colleagues recently demonstrate associations of the –1993C allele with aspirin- induced asthma in a Japanese case-control analysis. In that study, no association of this allele or three others was observed with either adult or childhood asthma not related to aspirin intolerance (8). No significant associations with –1993 were observed in our cohort, and we do not have detailed information regarding aspirin sensitivity in CAMP, preventing evaluation of the associations of –1993C with aspirin-induced asthma in our cohort. Our major finding was evidence of a relationship between multiple T-bet SNPs and AHR, a phenotype that was not evaluated in the other populations. We also found association with T-bet SNP c.–7947 and asthma. This variant is located approximately 8 kb upstream from the T-bet transcript and was not evaluated in the other studies. Because this variant is not in strong LD (in our cohort) with those SNPs that overlap between the four studies, it is unclear whether c.–7947 would have demonstrated association with asthma in the other populations. In addition, the strength of association between asthma and c.–7947 in our cohort is at best modest, and it is quite possible that this observation was observed by chance and is not biologically meaningful. Additional replication studies of this locus are warranted. One striking area of agreement across all studies is the high degree of sequence conservation across T-bet: only two nonsynonymous coding variants (His33Gln and Ile339Val) were found in the aggregate pool of 186 chromosomes surveyed in these four reports (48 Korean chromosomes, 46 Finnish chromosomes, 48 Japanese chromosomes, and 44 white chromosomes), and the bulk of observed variation mapped to regions of relatively low phylogenetic conservation. As suggested by Ylikoski (7), this restricted T-bet diversity likely reflects the important functions of T-bet for proper immune system development. As with the two previous studies that assessed associations with IgE levels, we did not find evidence of a relationship between T-bet SNPs and serum IgE levels in either CAMP or NAS. The lack of any association with atopy-related phenotypes in four populations suggests that T-bet genetic variation does not appear to directly influence clinical markers of atopy such as total serum IgE or blood eosinophil levels. Given our understanding of T-bet's role in T-cell differentiation, these results are perhaps surprising and the associations with AHR all the more intriguing: through what mechanisms could variation in T-bet alter airway responsiveness without affecting these atopic phenotypes? T-bet–deficient mice manifest AHR to methacholine, peribronchial eosinophilic inflammation, and chronic peribronchial fibrosis and remodeling (3), suggesting that the airway responsiveness–associated variants (or untyped variants in LD) somehow confer a T-bet–deficient state. This notion is further supported by the observations of reduced T-bet expression in CD4+ lymphocytes in asthmatic airways in humans (3). It has been shown that CD4+ lymphocytes in asthmatic airways have Th2-type cytokine profiles, even among patients with asthma who do not manifest classic atopic features (so-called intrinsic patients with asthma), suggesting that subtle Th2 skewing is present in most forms of asthma (29). These effects may be mediated by IL-13, which is expressed at higher levels in asthmatic bronchial mucosa (30) and mediates eosinophil recruitment, mucus hypersecretion, and AHR in murine models, even in the absence of other Th2 cytokines IL-4 and IL-5 (31, 32). In T-bet–deficient mice, local blockade of IL-13 by intranasal administration of IL-13 blocking antibody reduces airway inflammation, airway responsiveness, and subsequent airway remodeling, suggesting that IL-13 derived from T-bet–deficient activated CD4+ lymphocytes residing in the airways plays a central role in this process (33). Although it is conceivable that the functional consequences of human T-bet SNPs are restricted to the airway, it is more plausible that these variants have more subtle effects on Th1:Th2 balance that is not discernible by assessing clinical phenotypes such as serum IgE or blood eosinophil levels. Functional studies of these variants and studies correlating T-bet genotype with cytokine expression profiles would help resolve these questions. It is noteworthy that Akahoshi and colleagues recently provided in vitro data suggesting that the –1993C allele (i.e., the allele associated with aspirin-induced asthma) may confer increased T-bet expression (8).
In summary, we have demonstrated evidence of an association between T-bet polymorphism and the degree of AHR among white children with asthma, with modest evidence that these findings can be extended to other populations demonstrating airway responsiveness. Because associations were observed with SNPs that are not in tight LD with each other, it is unclear whether or not any of the identified SNPs are the truly functional variants, or whether these loci are in linkage disequilibrium with functional variants that have not been typed. It is also possible that the c.–7947 associations relate to effects on a neighboring gene, TBK1-binding protein 1, which is situated approximately 21 kb upstream from T-bet and has been implicated in tumor necrosis factor-
The authors thank all families for their enthusiastic participation in the CAMP Genetics Ancillary Study, supported by the National Heart, Lung, and Blood Institute (NHLBI), NO1-HR-16049. They also acknowledge the CAMP investigators and research team, supported by NHLBI, for collection of CAMP Genetic Ancillary Study data. All work on data from the CAMP Genetics Ancillary Study was conducted at the Channing Laboratory of the Brigham and Women's Hospital under appropriate CAMP policies and human subject protections. The authors thank Jody Senter Sylvia, Mel Hernandez, Sarah Tobias, Michael Hagar, and Maura Regan for their assistance with sample management and genotyping.
Supported by National Heart, Lung, and Blood Institute (NHLBI) NO1 HR16049, P50 HL67664, and T32 HL07427 and National Institute of Allergy and Infectious Disease AI31541. B.A.R. is a recipient of a Clinician Scientist Award from the Canadian Institutes of Health Research (MC1-40745) and a Mentored Clinician Scientist Award from the NHLBI (K08 HL074193). E.-S.H. is a recipient of an Arthritis Foundation Postdoctoral Fellowship. This article has an online supplement, which is accessible from this issue's table of contents at www.atsjournals.org Originally Published in Press as DOI: 10.1164/rccm.200503-505OC on September 22, 2005 Conflict of Interest Statement: B.A.R., E.-S.H, K.V.S., K.T., S.P., A.L., R.L., C.G., J.D.R., and D.S. do not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. E.K.S. received grant support and honoraria from GlaxoSmithKline for a study of COPD genetics. He also received a $500 speaker fee from Wyeth for a talk on COPD genetics. L.H.G. has equity in and is a paid consultant for Mannkind Corporation, a biopharmaceutical company focused on the development and commercialization of treatments for diseases, including cancer and autoimmune diseases, that owns the rights with Harvard University to the T-bet technology. S.T.W. received a grant for $900,065, Asthma Policy Study, from AstraZeneca from 1997 to 2003. He has been a coinvestigator on a grant from Boehringer Ingelheim to investigate a COPD natural history model that began in 2003. He has received no funds for his involvement in this project. He had been an advisor to the TENOR Study for Genentech and has received $5,000 for 2003–2004. He received a grant from Glaxo-Wellcome for $500,000 for genomic equipment from 2000 to 2003. He was a consultant for Roche Pharmaceuticals in 2000 and received no financial remuneration for this consultancy. Received in original form March 31, 2005; accepted in final form September 22, 2005
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
METHODS
2–3) (
/nuclear factor-
B signaling (
