Generating diversity in human glucocorticoid signaling through a racially diverse polymorphism in the beta isoform of the glucocorticoid receptor

Abstract

Alternative splicing of the human glucocorticoid receptor gene generates two isoforms, hGRα and hGRβ. hGRβ functions as a dominant-negative regulator of hGRα activity and but also has inherent transcriptional activity, collectively altering glucocorticoid sensitivity. Single-nucleotide polymorphisms in the 3' UTR of hGRβ have been associated with altered receptor protein expression, glucocorticoid sensitivity, and disease risk. Characterization of the hGRβ G3134T polymorphism has been limited to a relatively small, homogenous population. The objective of this study was to determine the prevalence of hGRβ G3134T in a diverse population and assess the association of hGRβ G3134T in this population with physiological outcomes. In a prospective cohort study, 3730 genetically diverse participants were genotyped for hGRβ G3134T and four common GR polymorphisms. A subset of these participants was evaluated for clinical and biochemical measurements. Immortalized human osteosarcoma cells (U-2 OS), stably transfected with wild-type or G3134T hGRβ, were evaluated for receptor expression, stability, and genome-wide gene expression. Glucocorticoid-mediated gene expression profiles were investigated in primary macrophages isolated from participants. In a racially diverse population, the minor allele frequency was 74% (50.7% heterozygous carriers and 23.3% homozygous minor allele), with a higher prevalence in Caucasian non-Hispanic participants. After adjusting for confounding variable, carriers of hGRβ G3134T were more likely to self-report allergies, have higher serum cortisol levels, and reduced cortisol suppression in response to low-dose dexamethasone. The presence of hGRβ G3134T in U-2 OS cells increased hGR mRNA stability and protein expression. Microarray analysis revealed that the presence of the hGRβ G3134T polymorphism uniquely altered gene expression profiles in U-2 OS cells and primary macrophages. hGRβ G3134T is significantly present in the study population and associated with race, self-reported disease, and serum levels of glucocorticoids. Underlying these health differences may be changes in gene expression driven by altered receptor stability.

Figure 1

Schematic representation of the glucocorticoid receptor (GR) gene and polymorphisms included in the study. Indicated are the functional domains of GR and the exons composing those domains. Polymorphisms are indicated within domain and exon. The nucleotide variant is listed for each polymorphism.

Figure 2

G3134T hGRβ polymorphism affects GR stability resulting in elevated GR protein expression. (a) Semi-log plot of relative mRNA expression vs time. RNA decay was measured by treating wild type (WT) hGRβ and G3134T hGRβ cells with 5 μg/μl of actinomycin D and harvesting mRNA at defined time points. Expression of hGRβ was measured by quantitative RT-PCR and values were normalized to the housekeeping gene Cyclophilin B (PPIB). mRNA expression for WT hGRβ at time 0 was set to 100% relative mRNA and expression at subsequent time points were set relative the expression at time 0. Data points show mean ± s.e.m. of four biological replicates and **P<0.01. (b) hGRβ protein levels were quantified by western blot analysis and normalized to levels of the housekeeping protein β-actin (n = 6–8 per genotype). The results represent the mean ± s.e.m. and *P<0.05. (c) hGRβ levels were measured by flow cytometry and the distribution of data are displayed for the % of cells that stained positive for hGRβ and the median fluorescence intensity (MFI) for hGRβ-positive staining cells. Patients with no copies of rs6189, rs6190, rs41423247, rs6191, and rs6198 were used for the 0 copy group (n = 10) and patients with two copies of rs6191 (G3134T hGRβ) and varying copy number of the other polymorphisms were used for the two copy group (n = 9). Lymphocytes and monocytes were determined from the total Peripheral blood mononuclear cell (PBMC) population by forward and side scatter. (d) Immunofluorescence of the glucocorticoid receptor was performed in fixed U-2 OS cells transfected with WT or G3134T hGRβ. Glucocorticoid receptor expression is shown in green and DRAQ5 staining of the nuclei is shown in red. (e) Immunofluorescence of the glucocorticoid receptor (green) was performed in U-2 OS cells stably expressing hGRα transfected with empty vector, WT hGRβ, or G3134T hGRβ and treated for 1 h with 100 nM Dex. Images taken at × 630.

Figure 3

Global gene expression is altered by G3134T hGRβ polymorphism. (a) mRNA isolated from three biological replicates of each genotype were analyzed by the Agilent Whole Human Genome 4 × 44 multiplex format oligo array for gene expression. Principal component analysis was performed in Partek Genomics Suite to visualize the variance within each genotype. The number of probes statistically different compared to U-OFF cells at P<0.05 with false discovery rate (step up) genotypes were sorted by Venn diagram. Heat map of significantly regulated gene probes by genotype was created with Partek Genomics Suite software and visually represent unique and commonly regulated targets. Red indicates induced genes and green indicates repressed genes. (b) All significantly regulated genes were analyzed by IPA software. A comparison analysis was performed, and the top 15 Diseases and Biological Functions and Canonical Signaling Pathways are reported. The activation score is a prediction of the activation state of the molecules in a given function. Orange bars predict an overall increase in the activity of the pathway, while blue bars indicate a prediction of an overall decrease in activity. (c) Genes significantly regulated in both genotypes (1319 total) were analyzed by IPA software. The top five Diseases and Biological Functions and Canonical Signaling Pathways are shown for each genotype. (d) A comparison analysis of genes uniquely regulated in either WT hGRβ cells or G3134T hGRβ cells was performed, and the top five Diseases and Biological Functions and Canonical Signaling Pathways are shown for each genotype. (e) Quantitative RT-PCR was performed on independent mRNA samples to validate expression changes in IL13RA2, CBLN2, TRAF1, and EDN3 mRNA from total isolated RNA. The reported fold change from microarray analysis is listed below the quantitative RT-PCR results. The results represent the mean ± s.e.m. and *P<0.05/**P<0.01.

Figure 4

Glucocorticoid-regulated gene expression is altered by the presence of the G3134T hGRβ polymorphism in human macrophages. (a) mRNA isolated from seven WT hGRβ participants and three homozygous G3134T hGRβ participants were analyzed by the Agilent Whole Human Genome 4 × 44 multiplex format oligo array for gene expression. Heat map of significantly regulated gene probes by genotype was created with Partek Genomics Suite software and visually represent unique and commonly regulated targets. Red indicates induced genes and green indicates repressed genes. (b) The 25 top Canonical Signaling Pathways were visualized and shared functions identified by lines connecting nodes. The intensity of the red color is proportional to significance of P-value as determined by the Fisher’s Exact Test. (c) The number of probes statistically different following 6 h treatment with 100 nM dexamethasone compared to vehicle at P<0.05 with false discovery rate (step up) were sorted by Venn diagram. (d) All significant glucocorticoid-regulated genes were analyzed by IPA software. A comparison analysis was performed, and the top 15 Diseases and Biological Functions and Canonical Signaling Pathways are reported. The activation score is a prediction of the activation state of the molecules in a given function. Orange bars predict an overall increase in the activity of the pathway, while blue bars indicate a prediction of an overall decrease in activity. (e) Gene significantly regulated by glucocorticoids in both WT hGRβ participants and homozygous G3134T hGRβ participants were analyzed by IPA software. The top five Diseases and Biological Functions and Canonical Signaling Pathways are shown for each genotype. (f) A comparison analysis of the genes uniquely regulated by glucocorticoids in either WT hGRβ participants or participants genotyped as homozygous G3134T hGRβ was performed, and the top five Diseases and Biological Functions and Canonical Signaling Pathways are shown for each genotype.