In our genomewide association study of more than 530,000 SNP markers with the use of a family-based screening algorithm, we identified 13 SNPs that were within the top 100 in statistical power for replication and associated with changes in FEV1
after the administration of inhaled glucocorticoids in the CAMP population. One GLCCI1
SNP, rs37972, was associated with the same phenotype with the use of the same genetic model in three of four independent populations of clinicaltrial populations with asthma, resulting in marked attenuation of the response to treatment with inhaled glucocorticoids (). This SNP is in complete linkage disequilibrium (i.e., perfectly correlated) with the functional rs37973, which down-regulates GLCCI1
expression — an effect augmented by exogenous glucocorticoids (, and Fig. 3 in the Supplementary Appendix
). In turn, glucocorticoid-induced GLCCI1
expression in B-cell lines was significantly associated with the response to inhaled glucocorticoids in the subjects with asthma from whom the cells were obtained. Thus, it is fully plausible that rs37973 causes a decremental response to inhaled glucocorticoids in patients with asthma through changes in GLCCI1
In addition to the functional role of rs37973, we noted significant associations between this variant and clinical responsiveness to inhaled glucocorticoids. In subjects who were homozygous for the wild-type allele, FEV1 improved by a factor of three after therapy with inhaled glucocorticoids, as compared with those who were homozygous for the mutant allele (pooled P = 0.0007). Moreover, subjects who were homozygous for the mutant allele were about two and a half times as likely to have a poor response to inhaled glucocorticoids (lowest quartile) as were those who were homozygous for the wild-type allele. Overall, this variant accounted for 6.6% of the variability related to the lowest-quartile response, indicating that variation at this locus may account for a substantial proportion of patients who have a poor response to therapy with inhaled glucocorticoids.
In this study, we identified a consistent genetic association that was replicated in three of four independent clinical trials with a combined total of 935 subjects. The family-based screening method used to identify the associated variant is ideally suited to pharmacogenetic studies in which sample sizes are limited and P values for true associations are likely to be modest. In our study, the statistical power was further enhanced by the use of repeated measures of the outcome variable during FBAT screening of the CAMP population.32
maps to chromosome 7p21.3 and contains eight exons. Its role in glucocorticoid signaling was first described by Chapman et al., who identified several differentially expressed sequenced tags in glucocorticoid-sensitive and glucocorticoid-resistant thymoma-derived cell lines after the administration of dexamethasone.37
One of these, GIG18
(glucocorticoid-induced gene 18), showed marked up-regulation. Now called GLCCI1
, it is expressed in both lung cells and immune cells, and its expression is significantly enhanced by the presence of glucocorticoids in asthmalike conditions (Fig. 2 in the Supplementary Appendix
may be an early marker of glucocorticoid-induced apoptosis.38
Apoptosis is a key mechanism through which glucocorticoids resolve lymphocytic and eosinophilic inflammation in asthma.39
Therefore, decreased GLCCI1
expression, as a result of rs37973, may reduce inflammatory-cell apoptosis, leading to a diminished clinical response to inhaled glucocorticoids.
Our study has several limitations. First, we specifically focused on 100 top-powered SNPs; the vast majority of SNPs in the genomewide association study were not tested. Clearly, a more comprehensive interrogation is warranted. Second, at enrollment, 85% of SOCS and SLIC participants reported recent use of inhaled glucocorticoids. Nonetheless, we noted a robust response to inhaled glucocorticoids during the common run-in period. Third, although our results would suggest that rs37973 has a functional role, other potential functional variants (e.g., alternative splice or intronic variants) may be present within GLCCI1. Although further mechanistic studies are needed to clarify the precise role of GLCCI1 in the response to asthma treatment, our elucidation of a functional variant to support the association finding far exceeds that of the typical genomewide study. Finally, owing to the small numbers of subjects in other racial groups, our analyses were restricted to whites, thereby limiting the generalizability of the results.
In conclusion, our genomewide association study of more than 530,000 SNPs revealed a novel functional SNP, rs37973, that decreases the expression of GLCCI1, a gene influencing the pharmacologic response to inhaled glucocorticoids in asthma.