Address correspondence and reprint requests to Ataúlfo Martínez-Torres, Departamento de Neurobiología Celular y Molecular, Universidad Nacional Autónoma de México, Instituto de Neurobiología, Campus Juriquilla, Querétaro, QRO 76230, Mexico. E-mail: firstname.lastname@example.org
γ-aminobutyric acid (GABA)ρ receptors regulate rapid synaptic ion currents in the axon end of retinal ON bipolar neurons, acting as a point of control along the visual pathway. In the GABAρ1 subunit knock out mouse, inhibition mediated by this receptor is totally eliminated, showing its role in neural transmission in retina. GABAρ1 mRNA is expressed in mouse retina after post-natal day 7, but little is known about its transcriptional regulation. To identify the GABAρ1 promoter, in silico analyses were performed and indicated that a 0.290-kb fragment, flanking the 5′-end of the GABAρ1 gene, includes putative transcription factor-binding sites, two Inr elements, and lacks a TATA-box. A rapid amplification of cDNA ends (RACE) assay showed three transcription start sites (TSS) clustered in the first exon. Luciferase reporter assays indicated that a 0.232-kb fragment upstream from the ATG is the minimal promoter in transfected cell lines and in vitro electroporated retinae. The second Inr and AP1 site are important to activate transcription in secretin tumor cells (STC-1) and retina. Finally, the 0.232-kb fragment drives green fluorescent protein (GFP) expression to the inner nuclear layer, where bipolar cells are present. This first work paves the way for further studies of molecular elements that control GABAρ1 transcription and regulate its expression during retinal development.
γ-aminobutyric acid (GABA) is the most abundant inhibitory neurotransmitter in the vertebrate brain and retina (Martin and Olsen 2000). GABA can activate two types of receptors: ionotropic GABAA and metabotropic GABAB receptors (Olsen and Sieghart 2008). These two receptors differ in their subunit composition and gating properties, and in their pharmacological and physiological profiles. GABAA receptors (GABAA-Rs) mediate fast inhibitory neurotransmission and allow the flow of chloride ions; they are assembled from five subunits encoded by a family of 19 genes including α (1–6), β (1–3), γ (1–3), δ, ε, θ, π, and ρ (1–3) subunits. GABAρ subunits assemble into functional, homomeric receptors (Martínez-Torres et al. 1998; Martínez-Torres and Miledi 1999; Watanabe et al. 2000; Simon et al. 2004; Olsen and Sieghart 2008) that do not desensitize during continuous application of GABA, are resistant to the GABAA receptor antagonist bicuculline and insensitive to the GABAB receptor agonist baclofen (Polenzani et al. 1991; Chebib 2004), and are activated specifically by the CACA and antagonized by the TPMPA (Ragozzino et al. 1996).
The expression profile of GABAA-Rs varies among brain regions in adult and neuronal types during development. GABAρ1 is found in several areas of the central nervous system (CNS), including the thalamus, hippocampus, pituitary, cortex, cerebellum, superior colliculus, spinal cord, amygdala, and neostriatum (Boue-Grabot et al. 1998; Wegelius et al. 1998; Enz and Cutting 1999; Rozzo et al. 2002; Mejia et al. 2008, Rosas-Arellano et al. 2011, 2012). GABAρ receptors have been widely studied as they serve as a major point of control along the visual pathway. In the GABAρ1 knock out mouse, the inhibition mediated by this receptor is totally abated in retina, demonstrating its role in proper neural transmission. GABAρ1 mRNA is very abundant in retina, and the receptor is located at the axonal terminus of bipolar cells, where it limits the release of glutamate onto ganglion cells (Qian and Dowling 1993; Pan and Lipton 1995; Lukasiewicz 1996; Koulen et al. 1997; Feigenspan and Bormann 1998; Fletcher et al. 1998; Ogurusu et al. 1999; McGillem et al. 2000; McCall et al. 2002; Sagdullaev et al. 2006).
GABAρ subunits are encoded by three genes; in mouse, GABAρ1 and GABAρ2 are arranged in tandem on chromosome 4, and separated by approximately 40 kb (Greka et al. 2000), and GABAρ3 is on chromosome 16. During mouse retinal development, GABAρ1 is highly expressed in ON bipolar cells after post-natal day 7 (Greka et al. 2000; Wu and Cutting 2001; Kim et al. 2008); nonetheless, transcriptional regulation of this gene is poorly understood. In silico studies indicate that a 5′-flanking region of the GABAρ1 gene lacks a TATA box and contains initiator elements (Inr) and several general transcription factor-binding sites (SP1 and AP1) that are well conserved in human, rat, and mouse.
Vertebrate GABAA-R genes are clustered and conserved with a basic structure with 9-coding exons (Tsang et al. 2007). Computational and experimental data revealed that GABAA-R promoters contain CpG islands, Inr elements, and a TATA-box. Core promoters for some GABAA genes have been located to 100- to 500-bp upstream of the translation start codon (For review see Steiger and Russek 2004; Joyce 2007). Nevertheless, there is no detailed information about the transcriptional regulation of GABAρ subunits.
Comprehension of the transcriptional regulation of GABAA genes can provide an understanding of the mechanisms that contribute to the etiology of neurological disorders and their responses to drug treatment, development, and maturing of the brain and retina. This first exploration of GABAρ1 subunit transcriptional regulation proposes that the mouse GABAρ1 receptor promoter is included within a 0.232-kb fragment upstream of this gene and that it activates transcription in the secretin tumor cells (STC-1) and cells in the inner nuclear layer (INL) of the retina, where bipolar cells are present.
Materials and methods
Animals and retina dissection
A colony of CD1 mice was bred and housed in the animal care facility at INB-UNAM. Newborn male mice were killed in accordance with protocols approved by the Institutional Animal Care and Use Committee of the UNAM. Eyes were removed and retinae were mechanically isolated. Retinae from three to five mice were pooled, frozen in dry ice, and stored at −80°C until use.
RNA isolation and qRT-PCR
Total RNA was isolated at least three times by processing 300–500 mg of retina using TRIzol® Reagent (Invitrogen, Carlsbad, CA, USA) following the manufacturer's instructions. Total RNA was suspended in diethylpyrocarbonate (DEPC)-treated H2O, and the concentration was measured by a spectrophotometer. The integrity of the RNA was assessed by electrophoresis in a 1% agarose gel.
Total RNA was reverse transcribed to cDNA by using SuperScriptTM II RNase H- Reverse Transcriptase (Invitrogen). One microgram of total RNA was mixed with 1 μL of 0.5 μg/μL oligo dT (Invitrogen), 1 μL 10 mM dNTP mixture, and DEPC-treated H2O in a final volume of 13 μL. The mixture was heated to 65°C for 5 min to denature the RNA and chilled on ice. Then, 4 μL 5X First-Strand Buffer, 2 μL 0.1 M dithiothreitol, and 1 μL SuperScript™ II were added to the reaction, which was incubated at 42°C for 50min. Reverse transcription was inactivated by incubation at 70°C for 15 min.
For qRT-PCR, the expression levels of GABAρ1 mRNA were compared and determined for P0 male mice retinae by means of qRT-PCR using the LightCycler™ (Roche Diagnostics, Indianapolis IN, USA). Primers were calibrated by using serial dilutions of cDNA to determine the relationship between cycle number (Ct) and expression of the GABAρ1 mRNA. Two independently synthesized samples were performed and amplifications were carried out in triplicate. ‘Light Cycler Fast Start DNA Master SYBR® Green I’ kit was used to perform these reactions, with actin and RplP0 genes as standards. Reactions included 5 μL of cDNA, 2.5 mM MgCl2, 250 μM sense and antisense primers (Table S1), 1 μL of SYBR® Green Taq ReadyMixTM, and 2 μL of water in a total reaction volume of 10 μL. Reaction conditions were 95°C for 10 min for one cycle (hot start), 40 cycles of 95°C for 10 s, 60°C for 10 s, and 72°C for 12 s. Nine microliters of each PCR reaction was electrophoresed in 2% agarose gels.
Two micrograms of total RNA isolated from adult male mice retinae was used to identify the transcription start sites (TSSs) using the GeneRacer™ kit from Invitrogen according the manufacturer's instructions. Total RNA was treated with enzymes to eliminate non-mRNA and truncated mRNA, and to remove end caps. The 5′-phosphate was ligated to the GeneRacer RNA oligo using the T4 RNA Ligase. Oligo dT was used for cDNA synthesis. PCR was performed with the rapid amplification of cDNA ends (RACE)2 primer (Table S2), with a denaturation step at 94°C for 1 min, followed by 33 cycles of 30 sec at 94°C, 30 s at 60°C, and 2 min at 68°C. The RACE1 primer (Table S2) was used for a nested PCR with the same conditions. The PCR products were purified and cloned into the pCR®4-TOPO® vector (Invitrogen). To detect the TSSs, 26 independent RACE clones were sequenced and aligned to the genomic sequence of GABAρ1 using the ClustalW2 software (EBI, Hinxton, Cambridge, UK).
For constructing the pBlue-TOPO®pGABAρ1 3 kb vector, the 3-kb fragment was amplified by PCR of genomic DNA isolated from CD1 mice and cloned into the pBlue-TOPO® vector (Invitrogen). The PCR product was purified using the illustra™ GFX™ PCR DNA and GEL Band Purification Kit (GE Healthcare Life Sciences, Piscataway, NJ, USA). The subsequent constructs were obtained by PCR using as template the 3-kb fragment. The PCR primers are presented in Table S3. For the luciferase constructs, products from BamHI and HindIII digestion were ligated into the pGL3 Basic vector (pGL-BV) at the BglII and HindIII sites. Deletions and mutants were generated by the Exsite PCR protocol with DpnI enzyme from the wild-type plasmid pGL3pGABAρ1 0.232 kb using the primers listed in Table S3. The 0.232-kb GABAρ1 promoter and the 0.232-kb GABAρ1 inverted promoter were subcloned from pGL3pGABAρ1 0.232 kb into the Sal/Sal sites of the CAG–GFP vector (provided by Dr. C.L. Cepko) to produce the 0.232-kb GABAρ1 promoter–GFP and the 0.232-kb GABAρ1 inverted promoter–green fluorescent protein (GFP) constructs. All constructs were confirmed by DNA sequencing. Mouse, rat, and human genomic DNA sequences were obtained from the NCBI genome database. DNA sequence analysis was performed using the MotifViz and TRANSFAC databases for predicting putative transcription factor-binding sites.
Cell cultures, transfections, and gene reporter assays
3T3, GT1-7, and HEK293 cell lines were grown in Dulbecco′s modified Eagle medium, (DMEM; Gibco, Grand Island, NY, USA) supplemented with 10% fetal bovine serum (FBS; Hyclone, Logan, UT, USA), penicillin 100 U/mL, and streptomycin 100 μg/mL (Gibco). STC-1 cells were maintained in DMEM supplemented with 15% horse serum, 2.5% FBS, 100 U/mL penicillin, and 100 μg/mL streptomycin. All cultures were maintained at 37°C and 5% CO2. All plasmids were purified using the Plasmid Midi Kit (Qiagen, Valencia, CA, USA). At least two independent plasmid preparations were used for transfection. One microgram plasmid DNA was cotransfected with 100 ng CMV promoter-driven Renilla luciferase construct (pRL-CMV) (Promega, Madison, WI, USA), which helped to normalize the transfection efficiency. Transfections were performed using 4.5 μL of Lipofectamine 2000 (Invitrogen) per well in six-well plates. Cells were harvested 48 h after transfection, and luciferase activity was measured using a luminometer (Turner Designs, Sunnyvale, CA, USA) with the Dual-Luciferase reporter assay system (Promega). The promoterless plasmid was the pGL3-basic vector (pGL-BV; Promega). The pGL3 plasmid, carrying the SV40 promoter, was the positive control to allow promoter strength comparison.
Dissected retinae of P0 male mice using 50% DMEM/50% F-12 (Gibco) were transferred to a micro electroporation chamber filled with a plasmid solution: 5 μg of pCAG-GFP as internal control for electroporation, 5 μg of pRL-CMV for normalization, and 10 μg of the constructs with the GABAρ1 promoter-driven firefly luciferase. For immunofluorescence, the pCAG-β-gal plasmid was used as internal control for electroporation. The electroporation protocol included five pulses of 25 V, 50 ms, each separated by 950 ms using a BTX ECM 360 electro square porator (BTX Instrument Division Harvard Apparatus, Holliston, MA, USA). Two retinae per plasmid were electroporated for each assay. Electroporated retinae were cultured on Nucleopore polycarbonate filters (0.2 mm pore size; Whatman, GE Healthcare Life Sciences, Piscataway, NJ, USA) at 37°C and 5% CO2 with 50% DMEM/50% F-12 (Gibco) supplemented with 10% fetal bovine serum (FBS; Hyclone), penicillin 100 U/mL, streptomycin 100 μg/mL (Gibco), and L-Glutamine (Invitrogen). Retinae were harvested 10–12 days after electroporation.
For organotypic cerebellar cultures, brains of P6 male mice were dissected and sliced (sagittal sections of 250 μm) in ice-cold low-sodium artificial cerebrospinal fluid medium (ACSF) containing 1 mM CaCl2, 10 mM D-Glucose, 4 mM KCl, 5 mM MgCl2, 26 mM sodium bicarbonate, 246 mM sucrose, and phenol red solution (1 : 1000), pH 7.3, previously aired with Carbogen (95% oxygen and 5% carbon dioxide). Slices were transferred to a micro electroporation chamber filled with a plasmid solution: 13 μg of pCAG-β-gal as internal control and 20 μg of the promoter constructs driving GFP expression. The electroporation protocol included five pulses of 85 V, 50 ms, each separated by 500 ms using a BTX ECM 360 electro square porator. Two slices per plasmid were electroporated for each assay. Electroporated slices were cultured on c30 mm culture plate inserts with 0.4-μm pores and 1.1 mL of culture media [75% Minimum Essential Medium Eagle (MEM), 25% heat-inactivated horse serum, 25 mM HEPES, 1 mM glutamine, 5 mg/mL glucose, penicillin (100 U/mL), and streptomycin (100 U/mL)] for 8 days.
Retinae were fixed in 4% paraformaldehyde for 30 min and then cryoprotected in 30% sucrose overnight at 4°C. Retinae were embedded in O. C. T. compound, frozen with dry ice, and stored at −20°C. Tissue was sectioned on a cryostat (20-μm cut cryosections). For immunostaining, slides were washed in 1X phosphate-buffered saline (PBS) for 5 min at 25°C and blocked with 5% goat serum in PBST (1X PBS and 0.1% Tween 20) for 30 min at 25°C. Primary antibodies were diluted in the blocking solution: chicken anti-GFP Abcam (Cambridge, MA, USA) ab13970 (1 : 500), rabbit anti-PKCalpha Sigma-Aldrich (St Louis, MO, USA) P4334-.2ML (1 : 500), and incubated overnight at 4°C in the dark. Antibody solution was removed, and the slides were washed in PBST thrice for 15 min each at 25°C. Secondary antibodies (from Jackson Immunoresearch, West Grove, PA, USA) diluted in the blocking solution were added: goat anti-chicken alexa488 103-545-155 (1.25 : 500) and goat anti-rabbit Cy3 111-165-144 (1 : 250), and incubated again overnight at 4°C in the dark. Slides were washed in PBST twice for 15 min each at 25°C and then counter-stained with propidium iodide or DAPI for 15 min. Slides were mounted using Vectashield H-100 (Vector Laboratories, Burlingame, CA, USA) and then imaged using a Zeiss LSM510 Meta confocal microscope (Carl Zeiss Advanced Imaging Microscopy, Germany).
Statistical analysis of data
qRT-PCR results were analyzed with the 2−ΔΔCt method and are shown as mean ± SEM. Significance testing is by one-way anova and Tukey′s test for group comparison using the GraphPad Prism 5 software (GraphPad Software Inc., La Joya, CA, USA). For transfection assays, samples were measured 7–12 times per duplicate; for transfection assays using deletions and mutants, samples were measured five times per duplicate; and for in vitro electroporated retinal explants, duplicate samples were measured four times. The values of luciferase activity were analyzed with Mann–Whitney U statistics using the GraphPad Prism 5 software. The means were compared considering a p-value ≤ 0.05 as a significant difference (mean ± SEM).
In silico analysis and identification of the transcription start sites for the GABAρ1 gene
Sequence analysis revealed an 88% identity between mouse and rat sequences and a 64% identity between mouse and human sequences; no TATA box was identified. The sequences predicted the presence of AP1, AP2, Ik-2, SP1, SRY, and E2F-myc transcription factor-binding sites and two putative Inr elements, all of which were well conserved in the three species (Fig. 1a). Sequence alignments with more species of vertebrates such as chimpanzee, rhesus, dog, opossum, marmoset, elephant, and chicken; showed considerable conservation in the 5′ flanking region of the GABAρ1 gene among these species (Figures S1 and S2).
We used total RNA from adult mouse retina to identify the TSS of the GABAρ1 receptor. RACE PCR products derived from full-length RNA were of different sizes, suggesting multiple TSSs for the GABAρ1 gene in mouse retina (Fig. 1b). Figure 1c shows three TSSs; the 159-bp TSS (which represents the 46% of the RACE clones), the 164-bp TSS (27%), and the 181-bp TSS (27%). We designated the first nucleotide A of the most abundant RACE clone (the 159-bp product) as the +1 position (Fig. 1d). Alignment of the 5′-RACE sequences to the gene mouse genomic Data Bank showed that the three TSSs are clustered in exon 1 without interruption, indicating that no intron was present at the 5′-end of GABAρ1 gene coding exon 1.
Identification of the minimal promoter region for the mouse GABAρ1 gene in STC-1 cells
To detect the GABAρ1 minimal promoter, we employed the luciferase reporter assay to analyze the ability of seven constructs to direct the expression of the reporter gene in four cell lines: STC-1, GT1-7, 3T3, and HEK. We isolated a 3-kb fragment of the 5′ upstream region of GABAρ1 gene and six serial deletions (Fig. 2a). Using the STC-1 cell line, which endogenously expresses the GABAρ1 mRNA (Jansen et al. 2000), the 3-kb, 2-kb, 1-kb, and 1-kb ∆506-bp constructs showed a 26% ± 4.60%, 34% ± 4.02%, 55% ± 4.10%, and 60% ± 3.41% of activity, respectively, as compared to the positive control (SV40 promoter, included to allow promoter strength comparison). The 1-kb fragment with the inverted orientation (1-kb*) showed only 8% ± 1.37% activity, suggesting that activity is orientation specific. The minimal promoter region, a 0.232-kb construct that contains all the potential transcriptional start sites and putative transcription factor-binding sites, presented the highest level of activity (161% ± 6.44%). Further deletions extended to 0.131-kb showed only basal activity compared to the control (8% ± 1.22%). The pGL3-basic vector (pGL-BV; Promega) was the negative control and showed 9 ± 1.69% activity. When we transfected all of these constructs into the GT-1, 3T3, and HEK 293 cell lines, we observed the same tendency, that is to say, serial removals of 5′ regions increased the transcriptional activity, and the 0.232-kb fragment showed the highest activity levels, indicating that this fragment corresponds to the basal promoter; the activity levels of the 1-kb* and the 0.131-kb constructs were not significantly different from the pGL-BV (negative control). Taken together, these results indicate that the minimal promoter region for driving expression of the GABAρ1 gene is located within the 0.232-kb region, which includes all the elements required to activate transcription in STC-1 cells.
cis-regulatory elements required for 0.232-kb promoter activity in STC-1 cells
To dissect the cis-regulatory elements that are critical for GABAρ1 promoter activity, we generated internal deletion mutants of the 0.232-kb construct and tested their activity for directing transcription in transfected STC-1 cells. The deleted transcription factor-binding sites were SP1, Inr-A, AP1, Inr-B, and SRY (Fig. 3a). Constructs with the deleted SP1, Inr-A, and SRY sites showed similar activity, with no significant difference as compared with the 0.232-kb fragment. Deletion of the AP1 site or the second Inr (Inr-B) drastically reduced the promoter activity by 47% ± 3.46% and 59% ± 2.90%, respectively (Fig. 3b). To confirm that the AP1- and Inr-B-binding sites contribute to the GABAρ1 promoter activity, we generated single-nucleotide mutants for the AP1 and Inr-B sites, as shown in Fig. 4a. Mutation of the AP1 site alone decreased the activity by 44% ± 3.92%, whereas the mutation of Inr-B alone reduced promoter activity by 55% ± 3.43%. Furthermore, when both sites were mutated, the activity levels were reduced by 85% ± 3.38%. These results indicate that the AP1 and the Inr-B sites are important for GABAρ1 promoter activity in transfected STC-1 cells.
In vitro electroporation does not modify endogenous expression levels of GABAρ1 mRNA in retinal explants
In preparation for exploring the promoter activity in mouse retinae, we performed quantitative RT-PCR to determine if electroporation modifies the GABAρ1 mRNA levels in mouse retinal explants (Fig. 5). We cultured P0 mouse retinae for 10 days after electroporation; the experimental groups included non-electroporated retinae (B), retinae electroporated with PBS 1X (C), and retinae electroporated with the CAG–GFP vector in PBS 1X (D). P10 mouse retinae were used as control (A). Actin and RplP0 genes were used to normalize the data. We did not observe a significant difference among groups (one-way anova p < 0.05). These results suggest that in vitro electroporation does not alter endogenous levels of GABAρ1 mRNA.
GABAρ1 promoter activity in retina
To determine whether the promoter activity of the 0.232-kb fragment extends to the retina, we electroporated in vitro P0 mouse retinae with several GABAρ1 promoter 5′ serial deletions used in previous experiments (2-kb, 1-kb, 1-kb*, and 0.232-kb fragments), and assayed the luciferase activity after 10 days in culture (Fig. 6a). Luciferase assays showed that the 2-kb and 1-kb fragments activate transcription in retina at the same level as the SV40 promoter (102% ± 16.23% and 128% ± 19.26%, respectively); the levels observed with the 1-kb* fragment and the pGL-BV were 20% ± 2.33% and 11% ± 3.07%, respectively. In contrast, the 0.232-kb fragment showed the highest level of transcriptional activity, an increase of 500% ± 16.50% relative to the SV40 control (Fig. 6b).
To validate the cis-regulatory elements that are critical for GABAρ1 promoter activity in transfected STC-1 cells, we electroporated the same internal deletion mutants of the 0.232-kb construct and tested their activity for directing transcription in retinal explants (Fig. 7a). Constructs with the deleted SP1, Inr-A, and SRY sites showed the same level of activity, with no significant difference as compared with the 0.232-kb fragment. Construct carrying the deletion of the AP1 site reduced its promoter activity by 56% ± 6.91%, and the second Inr (Inr-B) reduced the promoter activity by 48% ± 6.69% (Fig. 7b).
To test if the putative cis-regulatory elements are critical for GABAρ1 promoter activity, we electroporated in retina the 0.232-kb fragment and the versions carrying the mutated sites (Fig. 7c). Mutation of the AP1 site alone decreased the activity by 43% ± 8.77%, whereas mutation of Inr-B alone reduced promoter activity by 41% ± 9.53%. The activity level was reduced by 59% ± 6.12% when both sites were mutated (Fig. 7d).
To determine the neuronal cell layer of the retina where the 0.232-kb fragment activates transcription, immunofluorescence was detected after in vitro electroporation of mouse retinal explants using promoter construct-driven GFP expression with the 0.232-kb fragment and the 0.232-kb inverted fragment (Fig. 8). The CAG-promoter was used as a positive control and it activated transcription in all neuronal cell layers. Interestingly, the 0.232-kb fragment triggered transcription in the inner nuclear layer (INL) and ganglion cell layer (GCL), while the inverted version did not, suggesting that this is an orientation-specific region. A promoterless construct was used as a negative control (Fig. 8). To verify whether the GFP expression is present in ON bipolar cells, double immunofluorescence experiments were performed with an anti-PKCα antibody, a molecular marker for these cells (Haverkamp et al. 2003); as expected, GFP signal colocalized with the PKCα marker (Fig. 9). In sharp contrast, when we electroporated P6 mouse cerebellum slides with the 0.232-kb fragment, we could not detect significant levels of fluorescence after 8 days in vitro (Figure S3).
All these results indicate that the 0.232-kb fragment can activate specifically transcription in the INL, where ON bipolar cells are known to be present in retina; our results also indicate that this segment is orientation specific, and that AP1 and Inr-B sites are required for proper GABAρ1 promoter activity in mouse retina.
In this study, we analyzed the minimal promoter region of the GABAρ1 gene. The GABAρ1 receptor regulates rapid synaptic ion currents in the axon end of retinal ON bipolar neurons, controlling the excitatory input to ganglion neurons. GABAρ1 mRNA is highly expressed in retina, and the receptor is present in the terminals of the ON bipolar cells (Enz et al. 1996; Lukasiewicz 1996; Koulen et al. 1997; Boue-Grabot et al. 1998; Feigenspan and Bormann 1998; Fletcher et al. 1998; Wässle et al. 1998).
Initially, in silico analysis revealed that the GABAρ1 promoter lacks a TATA box, contains putative AP1, AP2, Ik-2, SP1, SRY, and E2F-myc transcription factor-binding sites, and two Inr elements that are well conserved in human, rat, and mouse. Indeed, the 5′-flanking region of the GABAρ1 gene is conserved among species of mammals and birds, but is quite dispersed when compared with amphibian and fish (Figures S1 and S2). Similar to many other genes, including genes coding for some GABAA-Rs, such as the human and mouse α3 (Mu and Burt 1999a), human α5 (Kim et al. 1997), human α6 (McLean et al. 2000), rat α6 (Jones et al. 1996; Bahn et al. 1997), mouse α6 (Jones et al. 1996; McLean et al. 2000), human β1 (Russek et al. 2000), human β3 (Kirkness and Fraser 1993), mouse γ2 (Mu and Burt 1999a, b), rat δ (Motejlek et al. 1994), and mouse δ (Sommer et al. 1990), the GABAρ1 gene promoter lacks a TATA box and contains multiple TSSs. Interestingly, the most abundant TSS detected by 5′ RACE contains the dinucleotide AC, which is a common transcription start point of RNA polymerase II transcripts (Breathnach and Chambon 1981). TATA-less promoters are not exclusive to ligand-gated ion channels. Genome-wide analyses of promoter regions have made it clear that TATA-driven, pre-initiation complex assembly is the exception in eukaryotic transcription, and only 10–20% of promoters contain a functional TATA box (Carninci et al. 2006; Yang et al. 2007). It has also been proposed that TATA-less promoters are commonly subject to evolutionary change in mammals (for review, see Sandelin et al. 2007).
Promoters cannot accurately be identified from sequence information alone, but they can be functionally classified as proximal or distal promoters. The proximal promoter is responsible for the correct positioning of the RNA polymerase II complex with respect to TSS, and its function is mediated by TATA and/or Inr elements that recruit the basal transcriptional machinery. On the other hand, the distal promoter contains multiple transcription factor-binding sites that confer greater specificity to transcription by stabilizing the pre-initiation complex (Lemon and Tjian 2000). After deleting the 5′ region of the GABAρ1 promoter, we observed that the minimal promoter, a 0.232-kb fragment, was enough to activate transcription in four different cell lines and in retina. Indeed, other GABAA subunits have minimal promoters with a comparable size; for example the rat α6 subunit gene, a 0.155-kb region (McLean et al. 2000); the human β1 subunit gene, a 0.279-kb fragment (Russek et al. 2000); and the human β3 subunit gene, a 0.143-kb minimal promoter region (Kirkness and Fraser 1993). This fact is noteworthy because it has been suggested that the GABAA-R clusters arose from an ancestral α-β subunit gene pair giving rise to the present GABAA-R subunits (Tsang et al. 2007). Remarkably, a model based on the selective action of putative scaffold/matrix attachment regions was proposed for the coordinate control and parallel expression of the α1 and β2 subunits, which are present in the mammalian GABAA-R gene cluster that comprises the α1, β2, γ2, and α6 subunits (Joyce 2007).
The mouse GABAρ1 minimal promoter contains a crucial Inr element, which is important for activating transcription, and is well conserved in the species of mammalian analyzed in our study. These elements are also included and required for functional activity of GABAA-R subunit promoters, such as the rat α5 (Kim et al. 1997); human, rat, and mouse α6 (Jones et al. 1996, 1996; Bahn et al. 1997; McLean et al. 2000, 2000), human β1 (Russek et al. 2000), mouse γ2 (Mu and Burt 1999a, b), and rat and mouse δ (Sommer et al. 1990; Motejlek et al. 1994). Deleting or mutating the second Inr element significantly decreased the reporter gene activity in transfected STC-1 cells and in electroporated retinal explants (Figs 3, 4 and 7). In TATA-less promoters, Inr elements regulate core promoter strength, determine the position of the TSS, and the interaction between Inr-binding proteins and components of the basal transcriptional machinery recruits RNA polymerase II to the transcription initiation complex (Weis and Reinberg 1992). Indeed, one Inr element is present in the 270-bp core promoter of the human GABAAβ1 subunit gene, and it mediates down-regulation of this gene (Russek et al. 2000).
Deletions and mutations of the AP1-binding site decreased significantly the GABAρ1 promoter activity. The AP1 transcription factors are a group of proteins that recognize and bind to specific AP1 DNA motifs in the promoter regions of genes; however, several other transcription factors recognize the AP1-binding site. The AP1 transcription factor is a dimer, and its complexity begins with the transcription factor itself (Morgan and Curran 1991). It is composed of many different combinations of hetero or homodimers, and the composition of AP1 determines which genes it regulates. Nevertheless, until now, it is unknown whether any member of the GABAA-R family interacts with the AP1 transcription factor and whether these proteins are important for transcriptional activity.
Several genes, such as the mouse metabotropic glutamate receptor type 6 (mGluR6), the Purkinje cell protein 2 (Pcp-2), and the Purkinje cell-specific L7 protein (L7) are expressed in retinal bipolar cells (Nordquist et al. 1988; Oberdick et al. 1988; Nawy and Jahr 1990; Shiells and Falk 1990; Berrebi et al. 1991; Nakajima et al. 1993; Nomura et al. 1994), but little is known about the transcriptional mechanisms that induce or repress their expression. Ueda et al. (1997) determined the spatial and temporal expression pattern of the mGluR6 using a transgenic mouse, which expresses the lacZ reporter gene under the regulation of a 9.5-kb fragment located at the 5′-flanking region of the mGluR6 gene. This 9.5-kb fragment triggers cell-specific and developmentally regulated expression of the mGluR6 gene in ON bipolar cells. Remarkably, the reporter gene (β-gal) and the endogenous mGluR6 gene are expressed in temporal coordination with the differentiation pattern of bipolar cells, indicating that this DNA fragment responds to a genetic program of retinal bipolar cell differentiation (Ueda et al. 1997). As far as we know, a detailed molecular dissection of the mGluR6 promoter and identification of putative transcription factor-binding sites have not been performed.
Pcp-2 is also expressed in retinal bipolar and cerebellar Purkinje neurons (Oberdick et al. 1990). The analysis of the upstream DNA sequence of the gene revealed the presence of general transcription factor-binding sites such as AP1, CRE, and Oct. Interestingly, a construct carrying 0.4-kb of upstream and 0.3-kb of downstream Pcp-2-flanking DNA to LacZ activated the β-galactosidase expression in a large number of neurons, including Purkinje neurons and bipolar cells. In contrast, a second construct carrying an additional 3.1-kb fragment of Pcp-2 upstream sequences limited the β-gal expression to Purkinje cells, whereas bipolar cells of the retina did not show activity of the reporter gene. These results suggest that the 3.1-kb fragment includes elements that suppress the transcription in bipolar cells, that is to say, additional components are involved in the negative regulation of the Pcp-2 gene within the retina (Vandaele et al. 1991).
For studying the L7 gene expression pattern, a transgenic mouse driving the β-gal expression was employed. This model contains 4-kb upstream of the start of the TSS and 2-kb downstream of the polyadenylation signal. High levels of the reporter gene were detected in retina and cerebellum; however, the essential sequences that drive the expression in bipolar neurons were not explored in this study (Oberdick et al. 1990).
When we tried to electroporate a construct carrying the 0.232-kb fragment in P6 cerebellum, we could not detect significant levels of GFP after 7 days in vitro (Figure S3). It is possible that other regulatory sequences are required for GABAρ1 expression in neurons and glia of the cerebellum, where this subunit has been found to be expressed (Harvey et al. 2006; Mejía et al. 2008, Reyes-Haro et al. submitted).
In conclusion, on the basis of all of these results, we suggest that a 0.232-kb fragment located upstream of the GABAρ1 gene is functionally important for activating transcription, and that the second Inr element (Inr-B) and the AP1 site are critical for promoter activity in transfected STC-1 cells and in vitro electroporated retinae. As the GABAρ1 receptor acts as a major point of control along the visual pathway, further studies should be considered to identify the specific transcription factors that participate in GABAρ1 gene transcriptional regulation during retinal development.
We thank I. A. Mártinez-Dávila, Ruíz-Alcibar, A. Gonzalez Gonzalez, and A. E. Espino-Saldaña for their technical assistance. We are in debt with Dr. C. L. Cepko for advices, reagents, plasmids, and facilities to perform some experiments in her laboratory at Harvard Medical School. Dr. M. M. Emerson for supervising dissection and electroporation of mice retina. Dr. A. Antaramián-Salas and A. González-Gallardo for DNA sequencing and qRT-PCR. Members of the Cepko/Tabin lab: A. Kan, Y. Chinchore, and T. Katz for advices and support. Dr. R. García-Villegas and Dr. T. E. Garay-Rojas for their advices and technical support. Dr. A. Varela-Echavarría, Dr. R. Salceda-Sacanelles, and members of their laboratories for their facilities and recommendations to this study. E. N. Hernández-Ríos for her support in confocal microscopy. We thank D. D. Pless for reading and editing the manuscript. All coauthors have seen and agree with the contents of the manuscript and there is no financial interest to report. This work was supported by grants from PAPIIT-UNAM IN202609 and IN205208 (AM-T and RM), and CONACyT (101851 to AM-T). AIM-P (210374) is recipient of fellowship from CONACyT, México. We thank the UNAM, INB, Programa de Doctorado en Ciencias Biomédicas, and the Shedid Fund.