Genotyping and Drug Response: Use of Single Nucleotide Polymorphisms (SNPs) versus Haplotypes to Predict Albuterol Efficacy

Note from the Editor

Suzanne Schubbert
University of California, San Francisco
Cancer Research Institute
2340 Sutter Street
Box 0128, Room S241
San Francisco, CA 94115-3024
sschubb@itsa.ucsf.edu

Drysdale et al. (1) and Martinez et al. (2) both report associations of genetic polymorphisms of the b2-adrenergic receptor with response to albuterol. However, the approach of these two groups differs significantly by genotyping either single nucleotide polymorphisms (SNPs) in isolation or multiple phased SNPs arranged on the same chromosome (haplotype). This has led to discordant results between the two studies. Drysdale et al. (1) suggest that the unique interactions of multiple SNPs within a haplotype, rather than any single SNP, can affect the bronchodilator response to albuterol in an individual. On the other hand, Martinez et al. (2) demonstrate that one polymorphism in amino acid residue 16 of b₂AR has a significant physiologic role in regulating responses to b₂AR agonists.

Drysdale et al. first identify 13 SNPs in a region spanning 1.6 kb of the b₂AR gene. They find that only 12 distinct haplotypes are represented in a population of four major ethnic groups, out of 8096 possible haplotypes representing all combinations of SNPs on a single chromosome. Moreover, only five of these haplotype pairs (one haplotype per chromosome) are common in an asthmatic cohort. Drysdale et al. (1) assess the bronchodilator response to albuterol in asthmatics to determine the relevance of the common haplotype pairs. Each haplotype pair is linked to clear differences in the in vivo response to b₂AR agonist, whereas there is no association between the response to albuterol and any individual SNP.

Martinez et al. (2) determine the distribution of two SNPs in a population of children and assess their response to b₂AR agonist. Gly-16 was previously shown to enhance agonist-induced downregulation of the b2 receptor (3,4). These authors find that homozygotes for Arg-16, when compared to Gly-16, are 5.3 times more likely to respond to albuterol than homozygotes; heterozygotes (Arg-16/Gly-16) were 2.3 times more likely to respond (2). This trend was observed for asthmatic and non-asthmatic children, and results were independent of baseline lung function, ethnic origin, and previous use of antiasthma medication.

Some of the specific findings of these two studies appear to be discordant. Martinez et al. (2) find that the Arg-16 allele is associated with a positive response to bronchodilators compared to the Gly-16 allele. In contrast, Drysdale (1) et al. find two haplotypes (haplotype 2 and 6) in their subject group that are associated with higher positive response, but both of these haplotypes have the Gly-16 polymorphism. Furthermore, Drysdale et al. (1) identify a haplotype that was associated with low bronchodilator response (haplotype 4); this haplotype contains the Arg-16 polymorphism. Martinez et al. (2) also notice the trend that homozygotes for Gln-27 tend to have higher positive responses compared to homozygotes for Glu-27, although not statistically significant. Yet, in the Drysdale study (1), haplotype 4 with Gln-27 is associated with decreased responsiveness and haplotype 2 with Glu-27 with increased responsiveness. In each case, the trends of the two studies go in opposite directions.

Differences in the experimental conditions used may account for discrepancies in the two papers. To begin with, the subject groups utilized in the two studies are different. Martinez et al. (2) genotype and assess bronchodilator response in 269 children. 188 of these subjects have two Caucasian parents, 40 have one Caucasian parent, and 41 have parents who are both Hispanic. Prevalence of asthma in these children, as defined by the authors, was 14.1%. Drysdale et al. (1) utilize a repository of multi-ethnic, "apparently normal" individuals to identify various SNPs in the b₂AR gene. This repository includes 23 Caucasian, 19 African-Americans, 20 Asians, and 15 Hispanic-Latinos. Then they use a cohort of asthmatic patients to determine whether the identified haplotypes of the b₂AR gene are associated with bronchodilatory response. The cohort consisted of 121 Caucasian patients. The two studies also employ slightly different methods to measure bronchodilator responsiveness. Martinez et al. assess responsiveness in the 269 children by administering two inhalations (180 mg) of albuterol using a metered dose inhaler and a spacer. Spirometry was performed before and 15 min after the albuterol doses using a standardized pneumotachygraphic method. Response to albuterol was expressed as % predicted post-bronchodilator forced expiratory volume in 1s (FEV1) - % predicted pre-bronchodilator FEV1. Drysdale et al. (1) performed spirometry on the asthmatic cohort before and 30 min after inhalation of 180 mg of albuterol delivered by nebulization. FEV1 and forced vital capacity (FVC) were determined in triplicate. The predicted values for these measurements were calculated based on standard algorithms. Similar to Martinez et al., Drysdale et al. considered the change in the percentage predicted FEV1 as the primary measure of responsiveness to albuterol.

I do not think that the more recent Drysdale study necessarily invalidates the Martinez study. The two studies examine different subject groups and report the findings they obtain from assessing these subject groups. Martinez et al. (2) based their study on previous clinical and biochemical work describing polymorphisms in the b₂AR gene (3,4). These authors focused on the question whether two different polymorphisms, at amino acids 16 and 27, were associated with differential responses to b₂AR agonists. Drysdale et al. (1) sought to assess the relevance of combinations of multiple SNPs, or specific haplotypes, in a span of the b₂AR gene for predicting the bronchodilator response to agonists.

Possibly, the discrepant results of the authors can at least in part be attributed to different subject groups. To begin with, Martinez et al. (1) utilize a population that exclusively consists of children. It is possible that the mechanisms by which b₂AR polymorphisms determine responsiveness to agonists in children may differ from those in adults. These children are of Caucasian, Hispanic, or mixed background. On the other hand, Drysdale et al. (2) assess bronchodilatory response in an asthmatic cohort of 121 Caucasian, asthmatic patients and identify the most common haplotype pairs present in the cohort. By assessing the predictive value of these haplotypes in response to albuterol the authors find that haplotype pair is significantly related to improvements in FEV1. In contrast to the Martinez results, no association was observed between the response to albuterol and any individual SNP. It is worth noting that these results are based on 121 Caucasian, asthmatic individuals. It is possible that different haplotypes are prevalent in other ethnic populations, which would confound any predictions among ethnic groups based on haplotype. In addition, perhaps results would differ if a mixed non-asthmatic/asthmatic population were studied. However, it should be noted that most of the subjects in the Martinez study were also Caucasian, and therefore, these discrepancies are not readily resolved in terms of different patient populations alone. Furthermore, Martinez observes similar trends for asthmatic and non-asthmatic children. However, similar trends may not be observed for asthmatic and non-asthmatic adults used in the Drysdale study.

Indeed, there remain discrepancies in the results of the studies performed by Drysdale et al. (1) and Martinez et al. (2) that cannot be readily resolved with the available information. Rather than assuming one study invalidates the other, both studies may have limitations imposed by the method of genotyping. Use of only one or two polymorphisms negates the possibility of significant interactions among SNPs in a given haplotype, whereas use of haplotypes may be confounded by variable prevalence among ethnic patient populations. It is prudent to examine carefully the subject groups and methods employed by the authors to better appreciate the strengths and weaknesses of the two studies and to evaluate the use of haplotype analysis in pharmacogenetic studies.

References:

1. Drysdale, C. M. et al. (2000) Complex promoter and coding region b2-adrenergic receptor haplotypes alter receptor expression and predict in vivo responsiveness. PNAS 97, 10483-10488

2. Martinez, F. D. et al. (1997) Association between genetic polymorphisms of the b2-adrenoceptor and response to albuterol in children with and without a history of wheezing. Journal of Clinical Investigation 100, 3184-3188

3. Green, S. A. et al. (1994) Amino-terminal polymorphisms of the human beta 2-adrenergic receptor impart distinct agonist-promoted regulatory properties. Biochemistry 33, 9414-9419

4. Reihsaus, E. et al. (1993) Mutations in the gene encoding for the beta 2-adrenergic receptor in normal and asthmatic subjects. Am. J. Respir. Cell Mol. Biol. 8, 334-349