Introduction

Summary Introduction Materials & Methods Results Discussion References

Catecholamines exert many of their important physiological functions by interacting with membranous -ARs. The activation of these receptors is transduced by the GTP-binding regulatory protein G_s into stimulation of adenylyl cyclase activity, which in turn increases intracellular levels of cAMP (1). -ARs are subdivided into three pharmacologically distinct but structurally related receptors (2-5). These receptors are products of separate genes and are unusual in that the ₁-AR and ₂-AR genes are intronless, whereas the ₃-AR gene contains a very short intron (2, 4, 6). The gene for the ₂-AR has been extensively studied including identification of the TSS, sequences involved in basal expression, and hormone-responsive elements for glucocorticoids and cAMP (3, 7-9). The 5-flanking regions for the human, nonhuman primate (rhesus macaque), rat, mouse, and ovine ₁-ARs have been cloned (10-15). Analogous to the ₂-AR, the early promoter regions of these genes are rich in GC sequences and lack a consensus TATA box. Several start sites for transcription were identified for the rat ₁-AR at ~250 and ~280 bp 5 to the ATG that initiate translation (12, 16). The TSS of the human ₁-AR was located 253 relative to the ATG (17), and the TSS for the human and rat ₂-ARs also occurred ~250 bp 5 to the ATG (3, 7). Therefore, similarities exist in the transcriptional initiation of ₁- and ₂-AR genes in humans and rodents. On the other hand, the TSS of ₁-AR genes in mouse and ovine species occurred at 415 and 660, respectively, demonstrating that there is variation among the species in the transcriptional initiation of ₁-AR genes (13, 14).

Another hallmark for the expression of the ₁-AR versus the ₂-AR genes is the tissue and developmental specificity of their expression. The ₁-AR gene is exclusively expressed in cerebral cortex, salivary glands, adipose cells, and VM, whereas the ₂-AR is present in lung, skin, skeletal muscle, and liver (18). Identification of promoter elements involved in basal and tissue-specific expression of these genes is an important avenue for characterization of this regulation. Another factor regulating the density of ₁- and ₂-ARs is their mRNA abundance in various tissues. The level of ₁-AR mRNA is inherently lower than that of ₂-AR mRNA in tissues expressing both subtypes of -AR (2). However, in tissues containing both mRNAs, the relative receptor abundance is not reflected in the mRNA concentration. For example, in primary heart cultures containing 80% myocytes (₁ expressors) and 20% nonmyocytes (₂ expressors), the ratio of ₁-AR to ₂-AR is 5:1 (19). However, 75% of the -AR mRNA represents the ₂ isoform, suggesting that the translational efficiencies of the ₁- and ₂-AR mRNAs may be different. Similarly, the density of ₁-AR mRNA in adipose tissue and brain is low despite the abundant expression of ₁-AR in these tissues (20). Therefore, both transcriptional and translational control mechanisms contribute to the abundance of ₁- and ₂-ARs.

In previous studies, the promoters of the rat ₁-AR and ovine ₁-AR were ligated to reporter genes and transfected into various cells lines (12, 21). With the use of these approaches, broad regions in each promoter that were involved in regulating basal transcription were identified. The goal of the current study was to identify more precisely the elements involved in appropriate basal expression of the rat ₁-AR gene. Using ₁-AR/luciferase chimeric genes in transient transfections, we identified elements responsible for the low level of ₁-AR transcription. These elements located between 3311 and 2741 and 125 and 100 function coordinately to decrease ₁-AR gene expression.

	Materials and Methods

Summary Introduction Materials & Methods Results Discussion References

Construction of chimeric luciferase reporter plasmids. A 9.8-kb EcoRI genomic fragment of the rat ₁-AR encoding 3.3 kb 5 to the initiator ATG, 1.5 kb of coding sequence, and 3.8 kb of the 3 region was excised from pUC118 with EcoRI (10). Digestion of the EcoRI genomic fragment with SacII generated a 3.2-kb fragment containing the entire 5-flanking region of the ₁-AR gene (GenBank accession numbers D00634 and U58651). The ₁-AR KpnI/SacII segment extending from 1251 to 126, relative to the initiator ATG, was subcloned into pBluescript SK⁺ and then excised with KpnI and SacI and cloned into pGL3basic (Promega, Madison, WI). pGL3basic is a promoterless luciferase expression plasmid containing a multiple cloning site 5 to the luciferase insert. To generate the 1251 to 100 pGL3basic plasmid, a 28-bp double-stranded oligomer (GGCTGCCCTGACCTGGCCGCGACCTCGC) encoding the region from 125 to 100 in the ₁-AR promoter and flanked with SacII-compatible ends was ligated into the SacII site. This insertion created the sequence of the ₁-AR promoter from 1251 to 100 driving the luciferase reporter gene. The 3311 to 126 and 3311 to 100 plasmids were generated by ligating the EcoRI/KpnI fragment extending from 3311 to 1250 5 to the KpnI site in the appropriate pGL3basic plasmid.

Cell transfections and luciferase assays. Ventricles were excised from 1-3-day-old rats, and VM were prepared and cultured as previously described (19). HepG2 and H4IIE cells were cultured in 3 ml of Dulbecco's modified Eagle's medium containing 5% fetal bovine serum, whereas SK-N-MC cells were cultured in medium containing 10% fetal bovine serum. Plasmids were introduced into these cells by calcium phosphate precipitation (22). HepG2 and H4IIE cells were seeded at a density of 4 × 10⁵ cells/60-mm dish, whereas VM and SK-N-MC cells were seeded at a density of 1.5 × 10⁶ cells/60-mm dish. After 24-48 hr, each dish was transfected with 10 µg of plasmid DNA composed of 5 µg of the smallest vector (pGL3basic), 3 µg of carrier pGEM7ZF⁺ DNA, and 2 µg of SV40/-galactosidase (pSV-gal; Promega) as a transfection control in a total volume of 1 ml. In all transfections, the amount of each ₁-AR/luciferase construct was increased to an equivalent molar ratio of pGL3basic, and the balance of the DNA was adjusted to 8 µg with pGEM7ZF⁺. Cells were exposed to the calcium phosphate precipitates for 16-20 hr, washed twice with phosphate-buffered saline, and recultured for an additional 48 hr. Cells were harvested in 250 µl of lysis buffer (25 mM Tris-phosphate, pH 7.8, 2 mM dithiothreitol, 2 mM EDTA, 10% glycerol, and 1% Triton X-100), and the lysates were clarified by centrifugation. Luciferase assays were performed using 20 µl of lysate and 100 µl of luciferase assay reagent (Promega), which were injected automatically into a Turner-20 luminometer. -Galactosidase assays were performed using 150 µl of lysate and the o-nitrophenyl--D galactopyranoside substrate as previously described (24). Protein was measured in 5 or 10 µl of extract using a detergent-compatible protein assay (BioRad, Hercules, CA).

Primer extension analysis. An antisense oligonucleotide denoted GLprimer2 (Promega), which is 5-CTTTATGTTTTTGGCGTCTTCCA-3 and corresponds to positions +111 to +89 relative to the multiple cloning site in pGL3basic (+2 and +24 in the luciferase gene), was 5 end-labeled with [-³²P]ATP using T4 polynucleotide kinase. RNA was prepared from VM that were transfected with chimeric ₁-AR/luciferase vectors (19). Ten micrograms of RNA and labeled primer was denatured for 10 min at 80° and then annealed for 45 min at 45°. The extension reaction was conducted at 42° using AMV Reverse Transcriptase for 20 min (Boehringer-Mannheim, Indianapolis, IN). The extension reaction was terminated by the addition of 5 µg of salmon sperm DNA and heat denaturation at 75° for 10 min followed by the addition of DNase-free RNase for 10 min at 37° (Boehringer-Mannheim). The extension reactions were treated with phenol/chloroform (1:1) and then ethanol-precipitated. The extension products (~300 bp) were denatured and analyzed on an 8% acrylamide sequencing gel containing 7 M urea. No extension products were observed on RNA extracted from VM that were not transfected with ₁-AR/luciferase vectors.

Preparation of rat heart nuclear extract. Left and right ventricles from 20 rats (21-24 days old) were cut into small pieces and suspended in 12 volumes of homogenization buffer composed of 10 mM HEPES, pH 7.6, 25 mM KCl, 2 M sucrose, 1 mM EDTA, 0.15 mM spermine, 0.5 mM spermidine, 10% glycerol, and 0.1 mM phenylmethylsulfonyl fluoride (27). The suspension was homogenized for 12 sec in a Brinkman Polytron (Brinkman Instruments, Westbury, NY) at setting 5 and homogenized further with six strokes using a Teflon/glass homogenizer. The homogenate was layered over a 10-ml cushion of homogenization buffer and centrifuged at 24,000 rpm for 30 min at 0° in an SW 27 rotor. The nuclear pellet at the bottom of each tube was resuspended in 2 volumes of nuclear lysis buffer composed of 10 mM HEPES, pH 7.6, 100 mM KCl, 3 mM MgCl₂, 0.1 mM EDTA, 1 mM dithiothreitol, 10% glycerol, and 0.1 mM phenylmethylsulfonyl fluoride and lysed by 20 strokes in a Teflon/glass homogenizer. The nuclear suspension was diluted to ~10 A₂₆₀ units/ml of nuclear lysis buffer, and a 0.1 volume of 4 M ammonium sulfate was added dropwise. The suspension was mixed by inversion at 4° for 30 min, and the viscous lysate was centrifuged at 50,000 rpm for 60 min in a Ti80 rotor at 0° to pellet chromatin. To the supernatant, 0.3 g/ml of solid ammonium sulfate was added, and the solution was mixed by inversion at 4° for 20 min. The suspension was centrifuged at 10,000 rpm in a swinging bucket JS-13 rotor for 40 min at 4°. The protein pellet was dissolved in 1 ml of dialysis buffer (25 mM HEPES, pH 7.6, 40 mM KCl, 0.1 mM EDTA, 1 mM dithiothreitol, 10% glycerol, and 0.1 mM phenylmethylsulfonyl fluoride) and dialyzed twice at 4° for 1.5 hr each time against 500 volumes of dialysis buffer. The nuclear extract was centrifuged at 14,000 rpm in a microfuge at 4°, and the supernatant was stored in small aliquots in liquid nitrogen. Rat liver nuclear extract was prepared according to the method of Gorski et al. (27).

Gel mobility assays. Double-stranded oligomers containing a 4-bp overhang were labeled with Klenow enzyme and [-³²P]dCTP (28). The binding reactions were performed at 0° for 20 min in a binding buffer composed of 80 mM KCl, 10 mM HEPES, pH 7.1, and 10% glycerol. Each binding mixture contained 1 µg of poly(dI-dC) as nonspecific competitor and proteins as indicated. The resulting complexes were resolved on 5% nondenaturing acrylamide gels in 25 mM Tris and 190 mM glycine at 4° (29). The oligomer derived from the ₁-AR sequence between 125 and 100 was CTAGAGGCTGCCCTGACCTGGCCGCGACCTCT (the underlined oligonucleotides represent those sequences in the ₁-AR gene). The relative concentration of double-stranded oligomers of the wild-type sequence between 125 and 100 and its mutants was determined by labeling 1 µl of oligomers annealed by Klenow enzyme and [-³²P]dCTP (28). Because 5 to 3 extension by Klenow polymerase occurs only on double-stranded oligonucleotides, the relative incorporation of the label into each oligomer indices the relative concentration of double-stranded DNA among the different samples.

DNase footprinting assays. The SacI genomic fragment encompassing nucleotides 3220 to +397 of the ₁-AR promoter was digested with PstI, and the 753-bp PstI fragment extending from 484 to +268 was subcloned into the PstI site in the multiple cloning region of pGEM3ZF⁺. The 753-bp PstI fragment was digested with NarI, and a 184-bp segment extending from 158 to +27 was cloned into the AccI site of pGEM3ZF⁺ plasmid. The NarI fragment containing plasmid was digested with HindIII, dephosphorylated, and labeled with -³²P-ATP and T4 polynucleotide kinase (28). The insert was released from the plasmid with BamHI and purified by electrophoresis on 1% agarose gels. The ³²P-labeled DNA fragment (20,000 cpm) was incubated in the absence or presence of nuclear extract (5-10 µg) and 1 µg of poly(dI-dC) in a binding buffer composed of 20 mM HEPES, pH 7.6, 0.1 mM EDTA, 1 mM dithiothreitol, 10% glycerol, and 50 mM NaCl for 30 min at 0° (30). Digestion with 0.03-0.1 unit of DNase I was allowed to proceed for 45 sec, followed by the addition of 150 mM NaCl, 0.7% sodium dodecyl sulfate, 15 mM EDTA, and 30 µg of yeast tRNA. The samples were extracted once with phenol-chloroform and ethanol-precipitated. Samples were resuspended in electrophoresis buffer and subjected to electrophoresis on 6% acrylamide in 8 M urea gels (30). The protected DNA sequences were identified from the autoradiogram of the gel. The sequence of the protected DNA was determined by running a separate lane containing a G sequence ladder generated according to the piperidine cleavage method of Maxam and Gilbert (31).

	Results

Summary Introduction Materials & Methods Results Discussion References

Transfection analysis identifies an element important for proper expression of the ₁-AR gene. Our initial experiments were designed to delineate regions in the ₁-AR promoter that are involved in regulating basal expression of the gene. The ₁-AR 5-flanking region contains three unique restriction sites: SacII site at 126 relative to the ATG initiating translation, KpnI site at 1251, and EcoRI site at 3311 (4). The KpnI/SacII and EcoRI/SacII fragments were linked to the reporter gene for firefly luciferase. Transient transfections were carried out in four cell lines from different tissues to assess absolute levels of expression as well as tissue-specific expression (Fig. 1). VM were prepared from ventricles of newborn rats by Percoll density gradient centrifugation and were composed of >90% myocytes that express ₁-AR (19). SK-N-MC cells are human neuroblastoma cells that express ₁- and ₃-ARs (32). HepG2 human hepatoma and H4IIE rat hepatoma cells are of liver origin and express ₂-AR (2, 33). Sequences extending from 1251 to 126 in the ₁-AR promoter were ligated into the promoterless pGL3basic vector and cotransfected with the pSV-gal vector into these cells. The relative amount of luciferase activity of the 1251 to 126 construct after correction for the transfection efficiency using -galactosidase was 52 ± 11 pg/mg of protein in HepG2 cells, 41 ± 9 pg/mg in VM, and 31 ± 10 pg/mg in SK-N-MC cells. Furthermore, the activity of this vector was ~40% of that of the SV40-driven pGL3control plasmid, in which the luciferase reporter gene is driven by the SV40 promoter/enhancer unit. The 1251 to 126 construct was expressed at high levels, which does not reflect the expected transcriptional activity for the ₁-AR gene.

View larger version (25K): [in this window] [in a new window]   Fig. 1.   Expression analysis of the rat 1-AR gene in mammalian cells. Left, sequences in the 5-flanking region of the 1-AR gene that were ligated into the promoterless pGL3basic vector. VM (Vent. Myocyte), HepG2 cells, and SK-N-MC cells were transiently transfected with 5 µg of equivalent of pGL3basic vector and 2 µg of SV40-gal as described in Materials and Methods. Cells were harvested at 48 hr after transfection and assayed for luciferase activity. Luciferase activity was corrected for -galactosidase to account for variations in transfection efficiency. Right, expression of the chimeric 1-AR/luciferase constructs relative to the expression of the SV40-driven luciferase plasmid pGL3control. Values are presented as mean ± standard error of three to six transfections. All transfections were performed in triplicate.

Characterization of the activity of the nucleotides between 125 and 100 in the ₁-AR promoter. We examined whether the region extending from 125 to 100 acts in either an orientation-specific or a position-specific manner to suppress transcription (Fig. 2A). When placed in its normal position in the promoter extending to 3311, the sequence from 125 to 100 inhibited the expression of the ₁-AR gene in both orientations. Relocation of this element from 126 into position 345 in the 5-flanking region of the ₁-AR restored full activity to the promoter, indicating that appropriate localization is important for the function of this element. Next, we tested whether the 125 to 100 sequence can inhibit the expression of a gene that uses a different set of regulatory proteins or factors (Fig. 2B). The PEPCK gene has a TATA box and is regulated by a variety of transcription factors (22, 23). The activity of the PEPCK-pGL2 construct in HepG2 cells was 38 pg/mg of protein. Insertion of the 125 to 100 sequence of the ₁-AR in front of the PEPCK 5-flanking region had no effect on PEPCK expression, indicating that inhibition by this element may be specific for the ₁-AR promoter.

View larger version (42K): [in this window] [in a new window]   Fig. 2.   Expression analysis of the nucleotides between 125 and 100 in regulation of 1-AR and PEPCK gene expression. Left, sequences in the 5-flanking region of the 1-AR and PEPCK genes. Arrow direction, orientation of the nucleotides between 125 and 100 in the promoter (i.e., , 5 to 3 orientation; , 3 to 5 orientation). Right, expression of the 1-AR and PEPCK constructs relative to the SV40-driven pGL2control vector. A, Expression of the 1-AR constructs relative to pGL2control in VM (Vent. Myocyte), HepG2 cells, and SK-N-MC cells was determined as described in the legend of Fig. 1 and in Materials and Methods. B, Expression of the 490 to + 73 sequence of the PEPCK promoter in the absence or presence of the nucleotides between 125 and 100 in the 1-AR 5-flanking region in HepG2 cells was determined. The activities of the pGL2control vectors in VM, HepG2 cells, and SK-N-MC cells were 31 ± 5, 39 ± 6, and 19 ± 4 pg/mg, respectively. Values are presented as mean ± standard error of three transfections. All transfections were performed in triplicate.

View larger version (33K): [in this window] [in a new window]   Fig. 3.   Identification of nucleotides between 125 and 100 that are involved in regulating 1-AR expression. a, Sequence of the nucleotides between 125 and 100 in the 5-flanking region of the rat 1-AR gene. Underlined, position of each 4-bp mutation. b, left, sequences that were ligated in front of the 3311 to 126 cassette of the 1-AR gene. These sequences were subcloned into the luciferase expression vector and transfected into VM (Vent. Myocyte), HepG2 cells, and SK-N-MC cells. Right, expression of the chimeric 1-AR/luciferase constructs relative to the expression of the SV40-driven luciferase plasmid pGL3control. Values are presented as mean ± standard error of three transfections. All transfections were performed in triplicate.

View this table: [in this window] [in a new window]   TABLE 1 Effect of mutating the 125 to 100 sequence on the expression of 1251 to 126 fragment of the 1-AR promoter The sequences between the nucleotides 1251 and 126 of the 1-AR 5-flanking region were cloned into the promoterless pGL3basic luciferase vector. The plasmid was cleaved at the SacII site, and the wild-type nucleotides between 126 and 100 (wt 125, 100) or the 125 to 100 sequence containing the base-pair changes outlined in Fig. 3 (Mut) were ligated into the vector. HepG2 and SK-N-MC cells were transiently transfected with 5 µg equivalent of pGL3basic vector and 2 µg of SV40-gal as described in Experimental Procedures. The activity of the SV-40 luciferase plasmid pGL3control in HepG2 and SK-N-MC cells was 150 ± 20 and 60 ± 9 pg/mg, respectively. The data represent the percentage expression of the 1-AR construct relative to the expression of the pGL3control plasmid. Values are mean ± standard error for the combined results of two transfections. Each transfection was performed in triplicate.

Transfection analysis of the progressively truncated 5-flanking region of the ₁-AR gene. To further characterize the interplay between the 3311 to 1251 region and the 125 to 100 domain in inhibition of the expression of the ₁-AR promoter, a series of 13 progressive truncations in the promoter between 3311 and 1251 were generated. These fragments were ligated to the ₁-AR promoter extending from the KpnI site at 1251 to 100 and tested for luciferase activity in transient transfection assays in VM and HepG2 cells (Fig. 4). Truncation of the ₁-AR 5-flanking region from 3311 to 3009 slightly increased the expression in VM. The activity of the ₁-AR promoter in HepG2 cells also increased modestly with each progressive truncation. Deletion of the bases between 2870 and 2740 significantly increased the expression of luciferase in VM and HepG2 cells. Excision of these 130 bases increased the expression by 3-fold in both cell types and boosted the expression to ~50% of pGL3control. Expression increased by an additional 2-fold in VM when the bases between 2740 and 2109 were excised. The activity of the ₁-AR promoter extending from 2109 to 100 in VM and HepG2 cells was 120% and 70%, respectively, relative to the SV40 luciferase vector. Continued deletion from 2109 to 1969 reduced the expression to ~30% in both cell types, suggesting the presence of a stimulatory element or elements in this region. The data indicate that the 5-flanking region between 3311 and 1251 contained negative and positive regulatory elements that exert profound effects on ₁-AR expression.

View larger version (35K): [in this window] [in a new window]   Fig. 4.   Expression analysis of progressively truncated 5-flanking regions of the 1-AR gene. Left, rat 1-AR 5-flanking sequences that were subcloned into the pGL3basic luciferase expression vector and transfected into VM (Vent. Myocyte) and HepG2 cells. Right, expression of the chimeric 1-AR/luciferase constructs relative to the expression of the pGL3control plasmid. Values are presented as mean ± standard error of three or four separate transfections. All transfections were performed in triplicate.

DNase I footprinting analysis of the early promoter region of the ₁-AR gene. The next experiments were designed to determine whether nuclear proteins could bind to any sites in the early promoter region of the ₁-AR gene. A 184-bp NarI fragment containing the sequences between 158 and +27 was labeled on the top strand, and the binding of rat heart nuclear proteins to this fragment was analyzed by DNase I footprinting (Fig. 5). Protected footprints were localized to the binding regions between 123 and 112 and between 106 and 100 (Fig. 5). These results indicate that the inhibitory domain identified by transient transfections can bind proteins in rat heart nuclei and suggest that two nuclear factors might be binding to this region. No other protected regions were identified on this strand. Domains D and C are continuously protected in the DNase I footprint. Domain A is also protected in the footprint, while the nucleotides in domain B demonstrate no ability to interact with nuclear proteins.

View larger version (42K): [in this window] [in a new window]   Fig. 5.   DNase I footprinting of 5-flanking regions in the 1-AR promoter. To perform DNase I footprinting, the 158 to +27 fragment was cloned into the AccI site in the polylinker of pGEM3ZF+. The polylinker was cleaved at the HindIII site, dephosphorylated, labeled with T4 polynucleotide kinase, and then cleaved at the EcoRI site. The labeled fragment was isolated by gel electrophoresis and incubated (20,000 cpm) without (lanes 2 and 3) or with (lanes 4 and 5) 0.5 or 6 µg of rat heart nuclear extract (RHNE) protein, respectively, followed by DNase I digestion with 15 munits (lane 3) or 30 munits (lanes 4 and 5) as described in Materials and Methods. The localization of the DNase I footprint (lane 5) was determined using Maxam-Gilbert G ladders (lane 1). The migration of the 72-bp DNA molecular weight standard is indicated in the gel .

Binding of rat heart nuclear proteins to the 125 to 100 sequence of the ₁-AR gene. Gel mobility assays were used to further characterize the interaction of rat heart nuclear proteins with the sequences between 125 and 100 (Fig. 6). To determine which nucleotides in the oligomer were required for binding of rat heart nuclear proteins, the wild-type oligomer as well as the oligomers containing the 4-bp mutations described in Fig. 3 were labeled and tested for their ability to bind nuclear proteins. As is shown in Fig. 6A, rat heart nuclear proteins formed two complexes, with the labeled wild-type oligomer containing the nucleotides between 125 and 100 (lane 2). The oligomer with a mutation in site D lost the ability to bind the more slowly migrating complex (lane 3). With the mutant C oligomer, a new complex was formed that migrated more quickly than the largest complex formed with the wild-type probe (lane 4). With mutant A, the more quickly migrating complex formed more strongly (lane 5). With the mutant B oligomer, the lower complex seemed to be migrating a little more quickly than the comparable band with the wild-type probe (lane 6). This observation suggests that these nucleotides might also contribute to the binding of nuclear proteins even though they are not protected in the DNase I footprinting assay. The data from the gel mobility assays confirmed the results found in the footprint assay: that at least two proteins can bind to this element. The binding of rat liver and rat heart nuclear extract was also compared (Fig. 6B). The liver nuclear extract forms only one complex with the wild-type oligomer (lane 8). Whether this protein represents the same or a different protein from the one in heart nuclear extract cannot be determined at this time.

View larger version (94K): [in this window] [in a new window]   Fig. 6.   Binding of nuclear proteins to the 125 to 100 element of the 1-AR gene. Procedures for the gel mobility assay and the preparation of proteins from rat heart nuclei extract (RHNE) are described in Materials and Methods. A, Binding reaction mixture contained 25,000 cpm of 32P-labeled oligomer representing the nucleotides between 125 and 100 in the 1-AR promoter. Lanes 2-6, 1 µl of rat heart nuclei extract (1 µg/µl) was added to each binding reaction. Lane 1, no protein was added to the mixture. Lanes 3-6, oligomers containing a 4-bp mutation as defined in the legend to Fig. 3 were labeled and used in the binding reactions. The binding reactions were conducted for 20 min at 4°. The protein/DNA complexes resolved on a 5% nondenaturing acrylamide gel at 4°. B, Binding of rat heart nuclei extract and binding of rat liver nuclear extract (RLNE) to the wild-type (WT) 125 to 100 1-AR oligomer were compared. The binding reactions and gel mobility assay were conducted as described above. The dried gels were exposed to a Kodak XAR-5 film overnight with one intensifying screen.

	Discussion

Summary Introduction Materials & Methods Results Discussion References

In this report, we present detailed analysis of elements that mediate the basal expression of the rat ₁-AR gene. The ₁-AR gene is expressed in a limited number of tissues and cell lines, including VM, rat C6 glioma cells, and SK-N-MC cells. Measurement of ₁-AR mRNA by DNA-excess solution hybridization in VM and rat glioma C6 cells revealed that ₁-AR mRNA is present at very low levels (19, 34). Our results indicate that basal expression of the ₁-AR gene is mediated by the interaction of two widely separated inhibitory domains. These elements are located ~120 bp and ~2800 bp 5 of the translational start site. Our data suggest that this unusual promoter architecture is responsible for the low basal activity of the ₁-AR gene and may contribute to tissue-specific expression of the ₁-AR.

A previous study by Searles et al. (12) indicated that the transcriptional activity of the rat ₁-AR promoter extending from 3354 to 1 in rat C6 glioma cells was ~30% of the thymidine kinase promoter activity. Deletion of the 5-flanking sequences from 3354 to 1064 increased the activity of these constructs to 70% of the thymidine kinase promoter. Activity of the promoter remained high with additional 5 deletions of the promoter. In our studies, expression of the 3-kb rat ₁-AR promoter in VM and SK-N-MC cells was also quite low: ~10% of that of the more active SV40 promoter/enhancer construct. Therefore, our data and those of Searles et al. (12) are in agreement and indicate that the transcription from the 3-kb ₁-AR rat promoter is quite low. The ovine ₁-AR promoter has been used in transient transfections into a variety of cell lines (21). In this promoter, a vector containing 2333 bp of 5 flanking sequence driving the CAT reporter gene was expressed a low levels, whereas a CAT vector containing 800 bp of the promoter was expressed at a 3-fold higher level. Like its rat counterpart, the ovine ₁-AR promoter has inhibitory sequences in the 5 end.

Using serial deletions of the ₁-AR promoter ligated to the luciferase reporter gene, we have expanded previous observations and identified several regions between 3311 and 2740 that are important for the suppression of rat ₁-AR transcription. Primarily, a region between 2870 and 2740 is responsible for transcriptional suppression. However, the function of the upstream inhibitory domains is absolutely reliant on the presence of an inhibitory element located 3 to the TSS. Removal of the 3 inhibitory element between 125 and 100 or relocation of this element to a site 5 to the TSS resulted in abundant expression in all cells that were analyzed, including hepatoma cell lines, which do not express ₁-AR mRNA or functional ₁-AR (2, 33). Deletion of the nucleotides between 125 and 100, however, did not alter the expression of constructs extending to 1251 (Fig. 1) or 484 (12). Therefore, the 125 and 100 element functions as a suppresser only in conjunction with other elements in the ₁-AR promoter and only when located 3 to the TSS.

The sequences between 125 and 100 normalize the activity of the 3-kb promoter in two ways. First, they suppress the expression of the ₁-AR/luciferase chimera to ~6-10% of the activity of the SV40/luciferase vectors. These levels of activity are within the expected parameters for a gene that is expressed at low levels (35). The concentration of ₁-AR mRNA in VM is ~0.25 amol/µg of RNA, which corresponds to ~28 molecules/cell (16, 19). These mRNA levels are 100-fold lower than those of actively transcribed genes such as the apolipoprotein E gene in the liver (36). Second, the sequences between 125 and 100 inhibit the expression of the ₁-AR/luciferase construct in HepG2 and H4IIE cells, which do not express ₁-AR mRNA (2). This observation suggests that this element can inhibit transcription in all cell types and that in conjunction with other sequences, it may contribute to the tissue-specific expression of the ₁-AR. Phenomenologically, the expression of the construct extending from 3311 to +100 seems to be similar to the expression of the ₁-AR in vivo. However, more detailed analyses of these constructs in vivo is required to verify that the tissue-specific expression of the ₁-AR promoter is under the regulation of this element.

In addition to the involvement of the 125 and 100 region in regulating basal expression, a second region that lies 3 to the TSS between nucleotides 186 to 211 was identified by Searles et al. as an inhibitor of basal expression (12). These experiments were conducted in the context of a luciferase vector extending from 1 to 484 relative to the start of translation. Our experiments did not evaluate the contribution of this element, and the probe used in our DNase I footprinting studies did not cover this region. Therefore, it is not known whether this element can bind nuclear proteins. This site is not conserved among the 5-flanking regions of the human, rat, and mouse ₁-AR genes, whereas the sequence in the 125 and 100 region of the rat ₁-AR gene is conserved among all the cloned ₁-AR genes (10-13). It will be interesting to evaluate the contribution of the 186 to 211 sequence in the context of the full 3.3-kb ₁-AR promoter.

Our results show that the 125 to 100 region can interact with two factors present in rat heart nuclear extract. We also found that the 125 to 100 element can bind the TR and mediate the thyroid hormone induction of the ₁-AR gene in VM.² Our data indicate that the complexes formed between rat heart nuclear extract and the 125 to 100 sequence in gel mobility assays do not contain the TR. However, during the preparation of nuclear extract, many nuclear factors are lost, so we cannot be certain that the factors binding to the 125 to 100 element in the gel mobility assays are binding in vivo. Our results indicate that this sequence is a multifunctional site that directs basal transcription and hormone responsiveness. It also raises the possibility that the TR could be one of the factors suppressing basal transcription of the ₁-AR gene because unliganded TR inhibits gene transcription (37). Future studies will be directed at identifying the factors that can bind the 125 to 100 element and characterizing their interactions with proteins on the more 5 region of the gene.