The gad2 promoter is a transcriptional target of estrogen receptor (ER)alpha and ER beta: a unifying hypothesis to explain diverse effects of estradiol

Abstract

Estradiol (E(2)) regulates a wide range of neural functions, many of which require activation of estrogen receptor alpha (ERalpha) and/or ERbeta, ligand-gated transcriptional regulators. Surprisingly, very few neural gene targets of ERs have been identified, and these cannot easily explain the myriad effects of E(2). GABA regulates most of the same neural functions as E(2), and GABAergic neurons throughout the brain contain ER. Therefore, we examined whether E(2) directly regulates expression of glutamic acid decarboxylase 2 (gad2), the enzyme primarily responsible for GABA synthesis for synaptic release. Using dual luciferase assays, we found that E(2), but not other gonadal steroids, stimulated the activity of a 2691 bp rat gad2 promoter reporter construct. Activation required either ERalpha or ERbeta, and ERbeta did not repress ERalpha-mediated transactivation. Site-directed mutagenesis studies identified three estrogen response elements (EREs) with cell-specific functions. An ERE at -711 upstream of the gad2 translational start site was essential for transactivation in both MCF-7 breast cancer cells and SN56.B5.G4 neural cells, but an ERE at -546 enhanced transcription only in neural cells. A third ERE at -1958 was inactive in neural cells but exerted potent transcriptional repression in E(2)-treated MCF-7 cells. Chromatin immunoprecipitation assays in mouse GABAergic N42 cells confirmed that E(2) induced ERalpha binding to a DNA fragment containing sequences corresponding to the -546 and -711 EREs of the rat promoter. Based on these data, we propose that direct transcriptional regulation of gad2 may explain, at least in part, the ability of E(2) to impact such a diverse array of neural functions.

Figure 1.

Diagram of relative locations of three putative EREs in rat gad2 promoter sequence used in these studies. Site locations are mapped based on the location of the translational start site (Trl).

Figure 2.

Effects of 10 nm E₂, T, P₄, or E₂ plus P₄ (E₂ pretreatment followed 6 h later with P₄) on rat gad2 promoter activity. gad2 promoter reporter construct was transiently transfected into MCF-7 cells, and cells were treated for 24 h with steroids after 18 h incubation in steroid-free media. Reporter activity was measured using a dual-luciferase assay with normalization to a renilla luciferase control. Each bar represents treatment mean ± SEM of at least nine samples. ***Significantly different from vehicle; p < 0.0001.

Figure 3.

E₂ activation of gad2 wild-type (WT) promoter and promoter constructs with mutated EREs. Activation was assessed using dual-luciferase assays. Designation of constructs is based on the positions of the mutated ERE upstream of the translational start site (see Materials and Methods). Constructs were transiently transfected into MCF-7 cells and treated with ethanol vehicle (Veh) or 10 nm E₂ for 24 h. Values are expressed relative to the wild-type construct in cells treated with vehicle. Each bar represents the mean ± SEM of values from at least nine independent samples. ***Significantly different from WT, vehicle control (p < 0.001).

Figure 4.

Effects of E₂ on reporter activity of a wild-type (WT) gad2 promoter construct or constructs with mutations of ERE sequences at −549, −711, and −1958 upstream of the gad2 translational start site (M-549, left column; M-711, middle column; and −1958, right column; for details, see Materials and Methods). Constructs were expressed in SN56.B5.G4 neural cells that were transiently transfected with empty vector (No ER, top row), hERα (second row), hERβ (third row), or both hERα and hERβ (ERαβ, bottom row). Cells were treated with 10 nm E₂ or ethanol vehicle (Veh) for 24 h before assaying reporter activity using dual-luciferase assays. The induction of promoter activity was expressed as fold activation compared with vehicle control. Each bar represents the mean ± SEM of values from at least nine independent samples. ^aSignificantly different from corresponding WT (p < 0.001); ^bsignificantly different from corresponding Veh control (p < 0.001); ^csignificantly different from corresponding Veh control (p < 0.01).

Figure 5.

Promoter sequence alignment for gad2 promoters of rat (AF090195), mouse (AB032757), dog (NW_876290), horse (NW_001867401), rhesus monkey (NW_001124107), chimpanzee (NW_001220649), and human (NT_008705AD2) using CLUSTALW (1.81) multiple alignment software. Designations of −1958, −711, and −546 refer to positions of putative EREs of rat GAD2 promoter relative to the translational start site. Shading of bases denotes identity with bases of the rat sequence, and asterisks indicate sites conserved among all species compared. BLAST analysis found insufficient homology to allow alignment between rodent and nonrodent species in the region containing the rat −1958 site.

Figure 6.

Results of ChIP using ERα antibody and DNA isolated from N42 cells. Cells were treated with ethanol vehicle (Veh) or E₂ for 2 h before they were harvested for ChIP assays. Samples were assayed using primers flanking the −532 ERE and the −693 ERE or a site with no ERE consensus sequences (for primer sequences, see Materials and Methods). Each data point represents the mean ± SEM of three independent samples, each analyzed in duplicate. **Significantly different from corresponding vehicle control; p < 0.01. Neg, Negative.