Common ADRB2 haplotypes derived from 26 polymorphic sites direct beta2-adrenergic receptor expression and regulation phenotypes

Abstract

Background: The beta2-adrenergic receptor (beta2AR) is expressed on numerous cell-types including airway smooth muscle cells and cardiomyocytes. Drugs (agonists or antagonists) acting at these receptors for treatment of asthma, chronic obstructive pulmonary disease, and heart failure show substantial interindividual variability in response. The ADRB2 gene is polymorphic in noncoding and coding regions, but virtually all ADRB2 association studies have utilized the two common nonsynonymous coding SNPs, often reaching discrepant conclusions.

Methodology/principal findings: We constructed the 8 common ADRB2 haplotypes derived from 26 polymorphisms in the promoter, 5'UTR, coding, and 3'UTR of the intronless ADRB2 gene. These were cloned into an expression construct lacking a vector-based promoter, so that beta2AR expression was driven by its promoter, and steady state expression could be modified by polymorphisms throughout ADRB2 within a haplotype. "Whole-gene" transfections were performed with COS-7 cells and revealed 4 haplotypes with increased cell surface beta2AR protein expression compared to the others. Agonist-promoted downregulation of beta2AR protein expression was also haplotype-dependent, and was found to be increased for 2 haplotypes. A phylogenetic tree of the haplotypes was derived and annotated by cellular phenotypes, revealing a pattern potentially driven by expression.

Conclusions/significance: Thus for obstructive lung disease, the initial bronchodilator response from intermittent administration of beta-agonist may be influenced by certain beta2AR haplotypes (expression phenotypes), while other haplotypes may influence tachyphylaxis during the response to chronic therapy (downregulation phenotypes). An ideal clinical outcome of high expression and less downregulation was found for two haplotypes. Haplotypes may also affect heart failure antagonist therapy, where beta2AR increase inotropy and are anti-apoptotic. The haplotype-specific expression and regulation phenotypes found in this transfection-based system suggest that the density of genetic information in the form of these haplotypes, or haplotype-clusters with similar phenotypes can potentially provide greater discrimination of phenotype in human disease and pharmacogenomic association studies.

Figure 1. β₂AR protein expression phenotypes of the common ADRB2 haplotypes.

COS-7 cells were transfected with constructs representing the β₂AR haplotypes shown in Table 1. Receptor protein expression was determined by quantitative radioligand binding. Results are from 7 experiments. *, p<0.05 vs. the other haplotypes.

Figure 2. Western blots confirm the radioligand binding method for ranking expression.

Monoclonal antibody to a non-polymorphic region of the β₂AR protein was utilized to confirm the radioligand binding method for ranking expression phenotypes of the β₂AR haplotypes. Shown is a single experiment revealing higher expression of haplotypes IV-1 and IV-3 compared to IV-4 and VI-2 β₂AR protein at the expected molecular weight (∼79 kDa). These results are consistent with those from radioligand binding of Figure 1. The control antibody was to Na⁺/K⁺-ATPase, a cell membrane protein.

Figure 3. ADRB2 mRNA levels of the transfected β₂AR haplotypes.

(A) mRNA levels for haplotypes IV-1 and IV-3 are higher, and I-1 lower, than levels of the other haplotypes. Results are from 6 experiments. *, p<0.05 vs. the other haplotypes. (B) relationship between ADRB2 mRNA levels and β₂AR protein expression. The r² for this relationship was 0.47. Two haplotypes (VI-1 and I-1) are greater than 1 standard deviation from the mean β₂AR/ADRB2 ratio.

Figure 4. Agonist-promoted downregulation phenotypes of the common ADRB2 haplotypes.

Transfected COS-7 cells were exposed to media alone or media with 10 µM isoproterenol for 48 hours. β₂AR protein expression was determined by quantitative radioligand binding. The results are from 7 experiments. *, % downregulation from the untreated state differs at p<0.05 vs. all other haplotypes.

Figure 5. Lack of a relationship between baseline β₂AR expression and the % downregulation.

The agonist-promoted downregulation, expressed as a percentage, was plotted against baseline β₂AR protein expression. There was no relationship (r² = 0.001, p>0.9) between the initial expression of β₂AR and the extent of downregulation. Results are from the 7 experiments of Figure 4.

Figure 6. Stratification matrix.

The phenotypes derived from the in vitro cell-based studies are stratified in a 2×2 matrix according to baseline β₂AR expression and the extent of agonist-promoted downregulation. Three bins were populated as shown and are denoted A, B, C for clarity.

Figure 7. Phylogenetic tree of β₂AR haplotypes with phenotypic annotation.

Shown is a phylogenetic tree constructed as indicated in Methods with the 8 β₂AR haplotype sequences. Shown are the expression phenotypes (I, increased; R, reference) and the dual phenotypes that incorporate expression and downregulation (denoted A, B, C from Figure 6). L1 and L2 represent the two ancestral lineages, while L2(1) and L2(2) are lineages from L2.