Cloning and Characterization of the Rat (cid:1) 1a -Adrenergic Receptor Gene Promoter DEMONSTRATION OF CELL SPECIFICITY AND REGULATION BY HYPOXIA*

Recent studies reveal important and distinct roles for cardiac (cid:1) 1a adrenergic receptors ( (cid:1) 1a ARs). Surprisingly, given their importance in myocardial ischemia/reperfu-sion, hypoxia, and hypertrophy as well as frequent use of rat cardiomyocyte model systems, the rat (cid:1) 1a AR gene promoter has never been characterized. Therefore, we isolated 3.9 kb of rat (cid:1) 1a AR 5 (cid:1) -untranslated region and 5 (cid:1) -regulatory sequences and identified multiple transcription initiation sites. One proximal (P1) and several clustered upstream distal promoters (P2, P3, and P4) were delineated. Sequences surrounding both proximal and distal promoters lack typical TATA or CCAAT boxes but contain cis -elements for multiple myocardium-rele-vant nuclear regulators including Sp1, GATA, and CREB, findings consistent with enhanced cardiac basal (cid:1) 1a AR expression seen in Northern blots and reporter constructs. Promoter analysis using deletion reporter constructs reveals, in addition to a powerful upstream enhancer, a key region (–558/ (cid:2) 542) important in regulating all (cid:1) 1a AR promoters with hypoxic stress. Gel shift analysis of this 14-bp region confirms a hypoxia-induced shift independent of direct hypoxia-inducible factor binding. Mutational analysis of this sequence identifies a novel 9-bp hypoxia To examine (cid:1) 1a AR promoter activity in different cellular contexts,

␣ 1 ARs 1 are members of the larger family of G protein-coupled receptors that mediate sympathetic nervous system responses such as smooth muscle contraction (1) and increased myocardial contractility (2,3); these responses occur predominantly via activation of phospholipase C-␤, resulting in the stimulation of both protein kinase C and inositol trisphosphate pathways (4). cDNAs encoding three ␣ 1 AR subtypes (␣ 1a , ␣ 1b , and ␣ 1d ) have been cloned and pharmacologically characterized in several expression systems, leading to the surprising finding that ␣ 1 AR subtype tissue distribution is species-dependent, with expression regulated at both gene and protein levels (4). To gain insight into transcriptional pathways governing ␣ 1 AR expression, several laboratories have initiated cloning and characterization of ␣ 1 AR subtype regulatory regions from different species, including the ␣ 1a AR (human (5), mouse (6)), ␣ 1b AR (mouse (7), rat (8), human (9)), and ␣ 1d AR (rat (10)).
Recent studies reveal an important and distinct role for ␣ 1a ARs in the heart, including enhanced contractility (2, 3) and hypertrophy (11)(12)(13)(14). Modulation of ␣ 1a AR expression occurs during pathological states, such as myocardial hypertrophy (13,14) and hypoxia (15), the latter usually secondary to ischemic injury. In hypertrophic models, classically performed in cultured neonatal rat cardiomyocytes, direct chronic ␣ 1a AR stimulation results in increased transcription of the ␣ 1a AR gene itself, providing a pathway for overall sustained ␣ 1 AR signaling in the heart (13,14). In contrast, both in vitro cardiomyocyte and whole animal models of hypoxia (1% O 2 , 72 h) reveal decreased ␣ 1 AR signaling, caused in part by reduced levels of ␣ 1a AR mRNA and protein (15). Interestingly, these studies also show that chronic hypoxia selectively attenuates ␣ 1 AR-mediated hypertrophy (15), suggesting that hypertrophy secondary to ischemic injury occurs via distinct, uncharacterized mechanisms. A number of pathways have been implicated in regulation of ␣ 1 AR signaling including protein kinase C, phosphoinositide 3-kinase, rho, ras, signal transducers and activators of transcription (STAT), c-Jun NH 2 -terminal kinase (JNK), p38, extracellular signal-regulated kinase (ERK), and calcineurin-dependent pathways (4). How these signals are integrated to regulate ␣ 1a AR expression also remains unknown.
Surprisingly, given the importance of ␣ 1a ARs in myocardial pathology and the use of rat neonatal cardiomyocytes to study hypertrophy, the rat ␣ 1a AR promoter has been neither isolated nor characterized. Therefore, to facilitate our mechanistic studies of myocardial ␣ 1a AR regulation, we cloned and characterized rat ␣ 1a AR 5Ј-UTR and regulatory sequences, investigated cell-specific expression, and characterized basal transcription in the presence and absence of the physiologic stress of hypoxia. Our data indicate that the rat ␣ 1a AR gene is both similar and yet distinct from mouse and human ␣ 1a AR genes. The rat ␣ 1a AR gene is transcribed from multiple promoters; a proximal promoter (Ϫ131 bp relative to ATG) as well as several clustered upstream distal promoters (P2, P3, P4, centered Ϫ2.1 kb relative to ATG). Both proximal and distal promoters are TATAless, containing consensus sequences for an initiator element as well as cis-elements for transcription factors Sp1, GATA, and CREB, shown to be important for cardiac gene transcription (16,17). These findings are consistent with enhanced cardiac basal ␣ 1a AR expression seen in Northern blots and reporter constructs. In addition to a powerful upstream enhancer, reporter gene and gel shift analysis reveals the presence of nuclear factor(s) able to associate with a 14-bp sequence of the rat ␣ 1a AR promoter in cardiomyocytes exposed to hypoxic stress. Systematic mutation of sequences within this region identified a potent novel 9-bp hypoxia response element (HRE). These findings provide an important foundation for elucidating specific ␣ 1 AR-mediated mechanistic pathways involved in distinct myocardial pathologies.

Library Screening
To isolate the 5Ј-UTR sequence of the rat ␣ 1a AR gene, a rat genomic library (DASH, average insert size 15 kb, Stratagene; La Jolla, CA) was screened using a liquid lysate approach. The ␣ 1a AR gene was isolated from phage DNA (18) using nested PCR amplification with two ␣ 1a AR-specific primers corresponding to residues 836 -860 and 776 -800 from the published cDNA sequence (19) (GenBank NM017191). T7 primer specific for the DASH vector polylinker was used as the opposing primer, creating a 4.2-kb PCR-generated genomic fragment. PCR fidelity was assured by subcloning and sequencing three independent amplification products.

Creation of Rat ␣ 1a AR 5Ј-UTR Deletion and Mutant Constructs
To determine the location of cis-elements important in conferring cell-specific and hypoxia-mediated regulation of the rat ␣ 1a AR gene, serial 5Ј-deletions of the full-length promoter fragment were made. A HindIII-NcoI digest of the 4.2-kb rat ␣ 1a AR fragment described above was cloned into SmaI-HindIII pGL2-Enhancer plasmid (Promega, Madison, WI) to generate the full-length rat ␣ 1a AR reporter construct. Serial deletions of the 5Ј-UTR of the rat ␣ 1a AR gene were created using either convenient restriction sites or a PCR-based approach; all PCR-generated constructs were sequenced to ensure fidelity. Site-directed mutagenesis was performed using the QuikChange XL kit (Stratagene) according to the manufacturer's recommendations. A 634-bp BamHI-NdeI fragment containing the desired mutation was sequenced and reintroduced into the original full-length reporter plasmid to ensure the integrity of all constructs.

Isolation of Total RNA
Total RNA was isolated by the Trizol method (Invitrogen). Genomic DNA contamination was removed by digestion of RNA samples with RNase-free DNase I (0.4 unit/l of RNA sample) for 30 min at 37°C, and degraded nucleic acid was removed using RNeasy spin columns (Qiagen; Valencia, CA). RNA concentrations were determined spectrophotometrically, and samples were stored at Ϫ80°C.

Northern Blot Analysis
A rat multitissue Northern blot containing 2 g of poly(A) ϩ mRNA in each lane was purchased from RNway laboratories (Seoul, Korea). Membranes were hybridized with a 32 P-labeled ␣ 1a AR RNA probe and washed as described previously (20). The blot was stripped and reprobed with a 220-bp GAPDH probe specific for nucleotides 663-855 of the published cDNA. Quantification was performed using ImageQuant software (Molecular Dynamics, Sunnyvale, CA), normalized to GAPDH levels.

Primer Extension Assay
Primer extension assays were used to corroborate the transcription initiation site results generated by RPAs. Primer extension oligonucleotide Primer 1 was made complementary to the region ϩ32/ϩ4, and Primer 2 corresponded to ϩ75/ϩ48, relative to the ATG. Oligonucleotides were end labeled with [␥-32 P]ATP using T4 polynucleotide kinase, and primer extension assays were performed as described previously (21) using 30 g of total RNA isolated from postnatal day 8 rat heart and isolated rat neonatal cardiomyocytes. Products from primer extension assays were size fractionated on a 6% polyacrylamide gel; a sequencing reaction using Primer 1 was used as a bp marker.
RT-PCR and Quantitation of ␣ 1a AR mRNA Levels-Because of the accuracy of competitive RT (cRT)-PCR over other RT-PCR methods and excellent target sensitivity, this method was used to quantitate ␣ 1a AR mRNA levels in several assays. Total RNA and competitor RNA were synchronously reverse transcribed and amplified using a commercially available kit (PerkinElmer Life Sciences). Control reactions with no RT were performed to ensure absence of contaminating genomic DNA. Samples were amplified as described (22). GAPDH was used to normalize mRNA levels using primers corresponding to residues 120 -137 and 338 -320 of the published cDNA sequence, respectively. Final PCR products were size fractionated on a 1.5% Tris borate-agarose gel bands, and normalized net band intensities were transformed into an amplification ratio and plotted logarithmically versus competitor amount. The x axis of the linear regression (least squares method) reflects the equivalence point at which the competitor concentration equals the initial mRNA concentration. This method provides the absolute amount of target transcript in starting samples.

General Cell Culture
To gain insight into mechanisms governing cell-specific ␣ 1a AR expression, rat1-fibroblasts, rat PC12 pheochromocytoma cells, and primary neonatal rat cardiomyocytes were used as model systems. Rat1fibroblasts were grown in Dulbecco's modified Eagle's medium (Invitrogen) supplemented with 10% fetal bovine serum, 100 units/ml penicillin/streptomycin, and 2 mM L-glutamine. PC12 cells (ATCC; Rockville, MD) were maintained in RPMI medium (Invitrogen) supplemented with 10% fetal bovine serum and 100 units/ml penicillin/streptomycin. Cells were plated onto 30-mm culture plates at a density of 200,000 cells/well in the appropriate medium.

Rat Myocyte Isolation and Culture
Postnatal (1-2 days postbirth) rat pup hearts were enzymatically dissociated with 73 units/ml collagenase and 0.6 mg/ml pancreatin, and cardiomyocytes were purified by Percoll density centrifugation as described previously (23). Myocytes were enriched further by preplating cells for 15 min on standard tissue culture dishes (Costar; Cambridge, MA) to remove fibroblasts. Cells were plated onto 30-mm laminincoated culture plates at a density of 500,000 cells/well and were maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum, 100 units/ml penicillin/streptomycin, and insulin/ transferrin/selenium (Invitrogen) and exposed to 90% ambient air and 10% CO 2 . Typical myocyte yields were 2 ϫ 10 6 cells/heart, with an average viability of Ͼ98% (determined by trypan blue exclusion).

Transient Transfection of ␣ 1a AR Reporter Constructs into Cultured Cells
To examine ␣ 1a AR promoter activity in different cellular contexts, ␣ 1a AR reporter constructs were transiently expressed in neonatal rat cardiomyocytes, PC12 cells, and rat1-fibroblasts followed by measurement of luciferase reporter gene activity. 24 h after initial plating, cells were washed with phosphate-buffered saline and fed with fresh medium. Cultures were transfected with 1 pmol of double CsCl-banded reporter plasmid and 0.5 pmol of ␤-galactosidase normalization plasmid (Promega) using the calcium phosphate precipitation method, as described previously (23). 24 post-transfection, cells washed twice with phosphate-buffered saline and maintained in appropriate medium for an additional 24 h. Transfected cells were harvested, and luciferase and ␤-galactosidase activity was assayed using the Dual-Light Assay System according to the manufacturer's recommendations (Tropix; Bedford, MA). Transfection efficiency was 5-10%, measured by visual ␤-galactosidase staining.

Hypoxia Exposure
Myocytes were placed in a hypoxia chamber and exposed to 88.5% N 2 , 10% CO 2 , and 1.5% O 2 for the lengths of time indicated in figure legends. Oxygen levels were monitored by a Fyrite Gas Analyzer (Illinois Scientific).

Preparation of Nuclear Extracts
Nuclear cell extracts were prepared from cardiomyocytes as described previously (24). Briefly, 1 ϫ 10 7 cells were washed twice with ice-cold phosphate-buffered saline and once with a solution containing 10 mM Tris-HCl, pH 7.8, 1.5 mM MgCl 2 , and 10 mM KCl supplemented with a protease inhibitor mixture containing 0.5 mM dithiothreitol, 0.4 mM phenylmethylsulfonyl fluoride, and 2 g/ml each leupeptin, pepstatin, and aprotinin (Sigma). Cells were lysed with a Dounce homogenizer, and nuclei were pelleted and resuspended in a solution containing 420 mM KCl, 20 mM Tris-HCl, pH 7.8, 1.5 mM MgCl 2 , and 20% glycerol, supplemented with the protease mixture described above and incubated at 4°C with gentle agitation. The extract was centrifuged at 10,000 ϫ g, and the supernatant was dialyzed twice against a solution of 20 mM Tris-HCl, pH 7.8, 100 mM KCl, 0.2 mM EDTA, and 20% glycerol. Protein concentration was determined using BCA reagent (Pierce) with bovine serum albumin as standard.

Electrophoretic Mobility Shift Assay (EMSA)
Oligonucleotide probe OL1, specific for region Ϫ573/Ϫ515, used for EMSA was 5Ј-GATC CTC CTC AAA TAT TTC AGA CCC ATG TCA CTT AGC CAG AAC TCC TAG ACG CTG GAG CTA GC-3Ј. The sense and antisense strands were annealed and end labeled with [ 32 P]ATP (PerkinElmer Life Sciences) using T4 polynucleotide kinase (Invitrogen). To measure DNA-protein interaction, 20 fmol (2-5 ϫ 10 4 cpm) of oligonucleotide probe was incubated with 5 g of nuclear extract and 0.1 g of sonicated, denatured salmon sperm DNA (Invitrogen) in 10 mM Tris-HCl, pH 7.8, 50 mM KCl, 50 mM NaCl, 1 mM MgCl 2 , 1 mM EDTA, 5 mM dithiothreitol, and 5% glycerol for 20 min at 4°C in a total volume of 20 l. The reaction mixture was size fractionated on 6% nondenaturing polyacrylamide gels at 4°C. Dried gels were subjected to autoradiography for 16 -48 h. For competition experiments, a 1,000-fold molar excess of unlabeled oligonucleotide was added to the binding reaction just before the addition of radiolabeled probe. Competitor oligonucleotides were 5Ј-CTCCTCAAATAT TTCAGA-3Ј, 5Ј-AGACCCAT-GTCACTTAG-3Ј, 5Ј-AGCCAGAACTCCTAGACG-3Ј, 5Ј-TAGACGCTG-FIG. 1. 5-Regulatory and 5-UTR DNA sequence from rat ␣ 1a AR gene. The rat genomic DNA sequence is shown with consensus sequences for transcription factor binding sites (highlighted in bold above the sequence). Where cis-elements overlap, the 5Ј-element is underlined and 3Ј-element in bold. TIS for P1, P2, and P3 are highlighted in bold and underlined; major initiation sites in P2 and P3 are labeled with arrowheads, and the P4 cluster is boxed. Initiation sites identified by primer extension are indicated by a bold vertical line above the nucleotide. A sequence encompassing P1 conforming to a consensus initiator element (Inr) is also boxed. The novel HRE is double underlined.

FIG. 2.
Identification of TIS of the rat ␣ 1a AR gene. A, schematic representation of RPA probes used for identification of the TIS. Restriction sites utilized for the generation of RPA Probes 1-9 are indicated. Primer locations for primer extension analysis are labeled and indicated by arrows. B, identification of the proximal promoter TIS by RPA. Total RNA from postnatal day 8 rat hearts (lanes 1-3) was hybridized to Probe 1 at 55°C. Yeast tRNA was used as a negative control (lane 4). The molecular weight marker (MW) was a 5Ј-end-labeled HaeIII digest of ⌽X174 bacteriophage. Four protected fragments were identified (arrows), with the major 125-bp product indicated by the wide arrow (n ϭ 3). C, primer extension assays were performed with antisense oligonucleotides Primer 1 (lanes 4 -6) and Primer 2 (lanes 1-3) using 40 g of total RNA from postnatal day 8 rat heart or cultured neonatal cardiomyocytes; tRNA served as a negative control (lanes 1 and 4). A DNA sequencing reaction (utilizing Primer 1 and rat ␣ 1a AR genomic DNA) is shown on the right for a bp marker. One major product was detected for each primer, with the difference between product lengths correlating precisely with their distance apart (⌬43 bp). Collectively, these data confirm a TIS at A(Ϫ131), relative to ATG. D, identification of the distal promoter TIS by RPA. Total RNA from 18-day gestation and 8-day postnatal rat hearts was hybridized to Probes 6 (lanes 5-8), 7 (lanes 1-4) and Probe 9 (lanes 9 -12) at 60°C. Unprotected probes along with fully protected Probe 6 are labeled. Protected products for P2 (Probe 6, lanes 5 and 6), P3 and the P4 cluster (Probe 7, lanes 2 and 3) are labeled (minor sites indicated by inverted v). The molecular weight marker (MW) is described in A above (n ϭ 3). Cyclophilin was used as a positive control probe, indicated by an arrow and labeled (inset). Yeast tRNA was used as a negative control for each probe (lanes 4, 7, and 12).
GAGCTAGC-3Ј, corresponding to sequential, overlapping 18-mers within the Ϫ573/Ϫ515 region, as described in the Fig. 6B legend. For gel supershift analysis, 1 l of monoclonal antibody against HIF-1␣ (StressGen Biotechnologies, Victoria, B. C., Canada) was added after the initial 20-min incubation, and the solution was incubated further for 10 min at 4°C before electrophoresis.

Statistical Analysis
Linear regression analysis of cRT-PCR data was calculated by the least squares method. Statistical analysis was performed using twotailed, unpaired Student's t tests with p Ͻ 0.05 considered significant. Results are presented as the means Ϯ S.E.

RESULTS
Cloning of the Rat ␣ 1a AR 5Ј-Regulatory Sequence and Identification of Proximal and Distal Promoters-To initiate studies on transcriptional regulation of the rat ␣ 1a AR gene, a 4.2-kb EcoRI-EagI fragment was subcloned from a genomic library and sequenced (3.9 kb of regulatory sequence is present upstream of the ATG shown in Fig. 1). To determine the location of 5Ј-regulatory sequence, the transcription initiation site (TIS) was identified by RPA using several overlapping probes that spanned the entire region, resulting in the identification of multiple promoters (RPA Probes 1-9; Fig. 2A). The proximal promoter was identified using Probe 1 ( Fig. 2A); four protected fragments are present (three minor, one major) ϳ85, 125 (major), and 140 bp in length (Fig. 2B). These products correspond to initiation sites of approximately Ϫ145, Ϫ130, and Ϫ85, respectively, relative to the ATG. Primer extension analysis was used to confirm RPA results (Fig. 2C). Two antisense oligonucleotide primers, corresponding to bp ϩ32/ϩ4 (Primer 1, relative to ATG) and ϩ77/ϩ50 (Primer 2) were utilized to map the TIS. In both cases, one major product was detected (163-bp product for Primer 1 and 208-bp product for Primer 2) corresponding to the 125-bp major product detected by RPA analysis. A bp sequencing ladder provides the exact site of transcription initiation as adenosine/-131 (CCAGCTG, relative to the ATG) in a region conforming to a loose initiator sequence (initiator; PyPyA (1)NT/APyPy (25)). Neither the shortest 85-/83bp, nor the longest 140-bp RPA products were detected by primer extension analysis. Because the major TIS RNA product was confirmed by two separate primers, it is possible that unconfirmed RPA products may be enzymatically generated because of inherent folding properties of the RPA probe used or that these potential TISs are below the threshold of detection by primer extension. We designate the main TIS located 131 bp upstream from the translation start site as the proximal (P1) ␣ 1a AR promoter.
Use of more distal overlapping RPA probes (Probes 5-9, Fig.  2A) revealed the presence of several closely clustered upstream promoters, with major start sites located at Ϫ2.2, Ϫ2.1 kb, and Ϫ1.9 kb, relative to ATG (Fig. 2D). No protected fragments were detected using Probe 9, indicating that the Ϫ2.2-kb cluster represents the most distal ␣ 1a AR transcription start points in heart tissue (Fig. 2D, right panel). These initiation sites were also confirmed using primer extension (data not shown), and the initiation sites are shown in Fig. 1. Thus, the rat ␣ 1a AR gene utilizes both proximal and distal promoters in heart located at Ϫ131 bp, Ϫ1.9 kb, Ϫ2.1 kb, and less abundantly at Ϫ2.2 kb, relative to the ATG; we designate these the proximal (P1) and distal (P2-P4) promoters.
Tissue-specific Expression of Rat ␣ 1a AR mRNA-As a first step in examining regulation of ␣ 1a AR gene expression, we determined the relative abundance of each transcript. Northern analysis was performed using mRNA isolated from rat heart and revealed two major products with molecular weights of ϳ3.5 and 1.7 kb (Fig. 3, lane 2). The difference in molecular weights of the two forms correlates well with the difference in length between the major products of the proximal and distal promoters (1.8 kb), corresponding to alternate promoter usage. Thus, the 1.7 kb product likely corresponds to P1-initiated transcripts. Similarly, the 3.5 kb band likely corresponds to the distal P2-P4 transcript cluster that comigrates closely because of poor resolution of products in this molecular weight range. Quantification of the two different forms shows that the higher molecular weight transcripts predominate, comprising 71% of the total ␣ 1a AR expression in rat heart. Interestingly, this transcript is expressed in heart and brain but is absent in other tissues examined (Fig. 3), suggesting that although P2-P4 promoters are important for cardiovascular and neural expression, P1 represents the key promoter in most cell types.
We next wished to determine absolute endogenous ␣ 1a AR mRNA levels in our model system (neonatal rat cardiomyocytes) using cRT-PCR. Results from cRT-PCR were compared with RPA results in a number of control cell lines and tissues (neonatal heart (8-day postnatal), adult heart, liver, kidney, rat1-fibroblasts, and PC12 pheochromocytoma cells) to ensure the fidelity of the technique (Fig. 4A). Both 8-day postnatal and adult rat heart have robust ␣ 1a AR expression; in contrast, rat kidney displayed much less ␣ 1a AR mRNA and PC12 cells even less (Table I). Rat liver and rat1-fibroblasts contain no detectable ␣ 1a AR mRNA. RPA results are highly consistent with cRT-PCR, with neonatal heart Ͼ adult heart Ͼ Ͼ adult kidney (Fig. 4B). Quantitation of ␣ 1a AR mRNA from these experiments demonstrates (Table I) a high degree of correlation between cRT-PCR and RPA methodologies.
Results show that rat ␣ 1a AR promoter activity is robust in cultured cardiomyocytes versus rat1-fibroblasts and PC12 cells, with promoter activity decreasing 38% in the latter two cells compared with cardiomyocytes (Fig. 4C, right panel). This is consistent with higher myocardial expression seen with the endogenous gene compared with PC12 or rat1-fibroblasts in which endogenous gene expression is either extremely low or absent (Fig. 4, A and B). Thus, although the latter two cells express little endogenous ␣ 1a AR mRNA, they do express factors capable of supporting high levels of ␣ 1a AR transcription, suggesting that post-transcriptional mechanisms may be in place to repress ␣ 1a AR mRNA expression in vivo.
To examine cell-specific expression of the proximal P1 promoter directly, 1.5 kb of proximal sequence was fused upstream of the luciferase reporter gene and transiently expressed in cell lines the same cell lines described above. Consistent with re- b, an RPA confirms quantitation of ␣ 1a AR mRNA levels by cRT-PCR. Total RNA (30 g) was hybridized to radiolabeled rat ␣ 1a AR RPA probe (upper panel) and cyclophilin control probe (lower panel). tRNA (lane 7) was used as a negative control. The results are representative of four independent experiments. c, cell-specific expression of rat proximal and distal ␣ 1a AR promoters. Either the full-length 3.9-kb ␣ 1a AR gene (containing both proximal and distal promoters) or a 1.5-kb P1 fragment was cloned upstream from the luciferase reporter gene and cotransfected with pSV-␤-galactosidase reference plasmid into indicated cell lines. Cell lysates were harvested and luciferase activity measured and normalized to ␤-galactosidase activity. Values are expressed as -fold change over basal expression of pGL2 enhancer plasmid (mean Ϯ S.E.; results are representative of three to five independent experiments).
sults from the full-length promoter fragment, the proximal promoter also directs higher levels of reporter gene expression in cardiomyocytes versus PC12 cells and rat1-fibroblasts, in which promoter activity decreases 61 and 69%, respectively (Fig. 4C, left panel). Together these results demonstrate that both proximal and distal promoter activity is consistent with high levels of cardiac basal expression seen with the endogenous ␣ 1a AR gene. FIG. 6. Basal expression of ␣ 1a AR mRNA under normoxic and hypoxic conditions. Full-length ␣ 1a AR 3.9 kb and deletion mutant reporter constructs were cotransfected with pSV-␤-galactosidase reference plasmid into rat neonatal cardiomyocytes. 6 h post-transfection, cells were exposed either to normoxic (23% O 2 ) or hypoxia (1.5% O 2 ) for 24 h; cell lysates were prepared and luciferase and ␤-galactosidase activities measured. Luciferase levels were normalized to ␤-galactosidase levels and expressed as -fold over basal pGL2 enhancer plasmid (mean Ϯ S.E.; results are representative of three or four independent experiments). On the left side of the schematic, shaded areas indicate 5Ј-UTR, and hatched areas represent a 5Ј-regulatory sequence. On the right side of the schematic, luciferase activities for normoxia (white bars) and hypoxia (black bars) are shown. b, luciferase activity of smaller deletion constructs under hypoxic stress was used to define the ␣ 1a AR HRE further. Two additional deletion constructs were created (Ϫ564, Ϫ374), and luciferase activity was measured and normalized to ␤-galactosidase activity as described above. *, p Ͻ 0.05 new deletion construct versus next longest reporter construct.
␣ 1a AR Promoter Activity under Normoxic and Hypoxic Conditions-To identify important regions governing ␣ 1a AR cardiac basal gene expression with normoxia and hypoxia, fulllength and serial 5Ј-deletion reporter constructs were transiently expressed into neonatal rat cardiomyocytes. As shown in Fig. 6, the full-length 3.9-kb reporter construct confers a 32-fold increase in luciferase levels over pGL2 empty vector under basal conditions in cardiomyocytes (Fig. 6A, white  bars). Deletion of the most distal 430 bp of the cloned promoter results in the loss of 95% of basal reporter gene activity (1.5fold over empty vector; p Ͻ 0.05) suggesting that a powerful enhancer is in the Ϫ3861/Ϫ3431 region. With further deletion to Ϫ3013 and Ϫ2100, reporter gene activity remain at low levels until the segment that contains P2 (Ϫ1926) is removed, where luciferase levels again increase in the Ϫ1538 construct to 3.6-fold over basal pGL2 enhancer vector. Removal of the Ϫ1538/Ϫ1233 region results in a 13% decrease in reporter gene levels (p Ͻ 0.05). Deletion to Ϫ950 results in a 33% increase in reporter activity (p Ͻ 0.01), consistent with the presence of basal repressor element(s) in this Ϫ1233/Ϫ950 region. Further deletion to Ϫ457 and Ϫ179 reduces ␣ 1a AR basal transcription to 27% and 47%, respectively (p Ͻ 0.01), relative to the fulllength fragment, suggesting the presence of independent enhancer elements in each of these latter regions, important for maximal ␣ 1a AR basal transcription. Finally, transcription occurs with as little as 48 bp of 5Ј-regulatory sequence upstream of P1 (Ϫ179 construct), suggesting that the proximal 48 bp are necessary and sufficient to confer promoter activity of the ␣ 1a AR gene, thus constituting a minimal promoter.
To identify cis-elements in the ␣ 1a AR promoter region conferring hypoxia-mediated decreases in ␣ 1a AR expression, we next characterized the full-length and deletion reporter constructs under hypoxic conditions (Fig. 6A, dark bars). Hypoxia induces a 59 Ϯ 8.2% reduction in transcriptional activity of the full-length reporter construct versus normoxic conditions (Fig.  6A compare dark versus white bars), correlating fairly well with the 77 Ϯ 10.6% repression of endogenous gene expression demonstrated previously in Fig. 5. Similar to results under normoxic (basal) conditions, deletion of the Ϫ3861/Ϫ3431 region results in a sharp reduction in ␣ 1a AR promoter activity under hypoxic conditions, consistent with the loss of the powerful basal enhancer in this region. Hypoxic transcription for subsequent deletion constructs parallels normoxic results, with hypoxic reporter gene activity largely remaining 30 -40% less than normoxic until P1 regulation is examined in the Ϫ520 FIG. 7. Binding of nuclear factors to rat ␣ 1a AR HRE in cardiomyocyte nuclear extracts exposed to hypoxic stress. A, isolated cardiomyocytes were cultured and exposed either to normoxic (20% O 2 ) or hypoxic (1.5% O 2 ) conditions for 24 h. Cells were then harvested and nuclear extracts prepared. 5 g of extract was used in each sample with 20 fmol of 5Ј-end labeled oligonucleotide (OL1) and incubated on ice for 20 min. Shifts were competed with either 1,000fold molar excess of self (S) oligonucleotide or nonspecific oligonucleotide (NS). HIF-1 in the binding complex was tested using a rabbit polyclonal antibody raised against HIF-1␣ (␣HIF). Samples were loaded on a 6% native polyacrylamide gel and resolved at 250 volts at 4°C. The gels were dried and exposed to autoradiography. Two specific shifts (1 and 2) are indicated by arrows. B, an EMSA was performed as described in A. Nuclear extracts from cardiomyocytes exposed to hypoxic stress (5 g) were incubated with 20 fmol of 5Ј-end-labeled OL1. Cold, competing oligonucleotides that encompass the Ϫ573/Ϫ515 region were added in either in a 20-fold or 500-fold molar excess. Cp1 is specific for Ϫ573/Ϫ555; Cp2, Ϫ558/Ϫ542; Cp3, Ϫ544/Ϫ528; and Cp4, Ϫ531/Ϫ515. construct. Deletion from Ϫ681 to Ϫ520 results in an 85% increase over the previous deletion construct, increasing reporter gene activity 4.1-fold over empty vector. Additional deletion to Ϫ457 again reduces transcription 77% (p Ͻ 0.01) versus the previous deletion construct with hypoxia, consistent with loss of an enhancer in this region. Finally, removal of sequence to Ϫ179 generates a construct with virtually identical activity under normoxic and hypoxic conditions, indicating creation of a constitutively active minimal promoter fragment. Deletion to ϩ88 abrogates all promoter activity, consistent with loss of all initiation sites (data not shown).
Because hypoxia decreases ␣ 1a AR expression in all reporter constructs (excluding the Ϫ520 construct), this suggests that both P1 and P2-P4 promoters may be affected by an element in this region. Therefore we investigated this further by creating smaller deletion constructs and analyzing them for transient reporter gene activity. These deletion constructs were designed either to remove an additional regulatory sequence or to mutate a consensus MCAT site at Ϫ512; MCAT is an element shown to be involved in basal regulation of the murine ␣ 1a AR gene (6). None of the smaller deletion constructs or the mutant MCAT construct had a statistically significant change in reporter gene levels under normoxic levels (data not shown), thus only data for reporter constructs is presented for hypoxic conditions (Fig. 6B, dark bars). Interestingly, mutation of the MCAT site did not have a significant effect on luciferase activity, suggesting differential regulation of the ␣ 1a AR gene in rat versus murine myocardium (data not shown). Removal of the ␣ 1a AR regulatory sequence to Ϫ564 had no significant affect on luciferase activity versus the Ϫ681 construct, whereas deletion to Ϫ374 results in a 54% increase in luciferase activity over the preceding reporter construct. Thus, deletion reporter gene analysis revealed the presence of three ␣ 1a AR HREs within the Ϫ564/Ϫ520, Ϫ520/Ϫ457, and Ϫ457/Ϫ374 regions. These regions were subsequently analyzed for direct protein binding by EMSAs.
Hypoxia-specific Binding of Nuclear Factors to an ␣ 1a AR HRE-Transient reporter expression experiments indicate that the Ϫ564/Ϫ374 region is important for hypoxia-mediated regulation of the proximal promoter, thus we wished to determine whether soluble nuclear factor(s) are able to bind specifically to this sequence. Oligonucleotide probes were generated (Fig. 7A, OL1, OL2, and OL3) which encompassed this region and subjected to gel shift analysis using normoxic and hypoxic extracts to determine whether a correlation exists between soluble nuclear factor binding and promoter activity. Only OL1, specific for region Ϫ573/Ϫ515, generated a new, specific complex induced under hypoxic versus normoxic conditions in extracts prepared from cardiomyocytes (Fig. 7A, lanes 6 -9, shift  1). This shift is competed by excess unlabeled self-oligonucleotide (Fig. 7A, lane 7) but is not competed by a nonspecific oligonucleotide (Fig. 7A, lane 8). Additionally, a specific constitutive shift is present in normoxic extracts (Fig. 7A, lanes 2-5,  shift 2), which increases with low oxygen stress. Again, this shift is competed by self-oligonucleotide (Fig. 7A, lane 3) but unaffected by a nonspecific oligonucleotide (Fig. 7A, lane 4). Finally, although no consensus sequence is present in the rat ␣ 1a AR promoter for the hypoxic transcriptional regulator HIF-1 (TACGTGCT), potential direct binding was examined. None of the shifts was affected by polyclonal antibody to HIF-1␣, suggesting that HIF is not a component of the complex that directly binds to this region of the rat ␣ 1a AR regulatory sequence (Fig. 7A, lanes 5 and 9). Together, these results suggest the presence of soluble nuclear factors that are able to associate with the Ϫ573/Ϫ515 region to confer hypoxia-mediated responsiveness to the rat ␣ 1a AR proximal promoter.
To narrow the ␣ 1a AR HRE further, the specific hypoxia shift was competed with even smaller overlapping oligonucleotides within the Ϫ573/Ϫ515 region (Fig. 7B). Only Cp2 was able to compete the binding in a dose-dependent manner, suggesting that protein binding occurs within the Ϫ558/Ϫ542 region of the rat ␣ 1a AR promoter (Fig. 7B, lanes 5 and 6). Interestingly, the highest concentration of Cp4 was able to compete to a lesser degree than Cp2, suggesting that proteins within the hypoxiabinding complex may be making contacts with the DNA in the Ϫ531/Ϫ515 region as well (Fig. 7B, lanes 9). Together, EMSA data suggest that nuclear factors associate with the Ϫ558/Ϫ542 and Ϫ531/Ϫ515 regions of the rat ␣ 1a AR promoter and modulate transcription under both constitutive and hypoxic conditions.
To determine whether the regions identified by EMSA were functionally relevant, these regions were mutated by site-directed mutagenesis in the full-length 3.9-kb reporter construct and transfected into cardiomyocytes under both normoxic and hypoxic conditions (Fig. 8B). Under normoxic conditions, only mutant 1 had a statistically significant affect on transcriptional activity, decreasing reporter gene levels 26% (Fig. 8B; p Ͻ 0.05). Importantly, under hypoxic conditions two sequential mutations within the Ϫ558/Ϫ542 region resulted in a marked attenuation of hypoxia responsiveness. Mutants 1 and 4 had little effect on hypoxia-mediated repression of reporter gene activity, but mutants 2 and 3 were able to increase ␣ 1a AR promoter responsiveness 1.9-fold versus the wild-type construct under hypoxic conditions. Thus, compared with the wildtype fragment under normoxic conditions, mutants 2 and 3 resulted in an attenuated 27 and 28% hypoxic decrease in reporter gene activity, respectively, versus the 59% decrease for the wild-type fragment (Fig. 8B; p Ͻ 0.05). Thus, the 9-bp element (Ϫ555 CCCATGTCA Ϫ547) is able to modulate ␣ 1a AR FIG. 8. A novel HRE regulates both proximal and distal ␣ 1a AR promoters in response to hypoxic stress. A, site-directed mutagenesis was used alter the full-length 3.9-kb reporter construct to create four independent reporter gene constructs (m1-m4), each containing a 3-bp transition mutation in the Ϫ558/Ϫ542 region. Mutated bp are indicated in bold and underlined. The sequence of all HRE mutants was confirmed by sequence analysis. B, wild-type and mutant reporter constructs were transiently expressed in neonatal cardiomyocytes as described in Fig. 5. **, p Յ 0.05 hypoxia versus full-length construct under hypoxic conditions (dark bars). responsiveness to hypoxia, although having little affect under normoxic, basal conditions. Coupled with evidence of a soluble nuclear factor binding to this region, inducible with hypoxia, this is strong evidence that this element is a potent HRE regulating the ␣ 1a AR gene under conditions of low oxygen stress.

DISCUSSION
Stimulation of cardiac ␣ 1a ARs has been shown to play an important role in myocardial hypertrophy (13,14), contractility (3), arrhythmias (26), and ischemic preconditioning (27). In the present study, we report molecular cloning and functional characterization of the rat ␣ 1a AR gene, including identification of multiple transcription initiation sites, with transcription initiation sites located proximally at 131 bp and a distal (upstream) cluster located 1.9, 2.1, and 2.2 kb upstream from the ATG. In addition to identification of a powerful enhancer/control element Ϸ3.6 kb upstream from the ATG necessary for optimal basal gene expression, we have demonstrated cell-specific transcription of both the proximal and distal promoters. Promoter activity is highest in cardiomyocytes versus other cell types examined, consistent with enhanced cardiac expression of the endogenous gene. Further, we demonstrate a correlation between endogenous gene levels and promoter activity with hypoxic stress and identify specific regions of the rat ␣ 1a AR gene capable of modifying basal and hypoxia-induced regulation of heterologous reporter constructs. This region appears to modulate hypoxia-induced decreases in expression for all promoters. Results from site-directed mutagenesis experiments support in vitro binding data and reveal a novel 9-bp HRE located 555 bp upstream from the ATG as a potent regulator of ␣ 1a AR transcription with hypoxic stress. These new findings provide clues to mechanisms underlying modulation of myocardial gene expression with disease and stress.
Northern blot analysis confirms that there are at least two transcripts in rat heart, with molecular weights of ϳ3.5 and ϳ1.7 kb. Using 5Ј-oriented probes, we saw no evidence of the higher molecular weight 9.5-and 11-kb transcripts reported previously to be expressed in rat heart (19). Because the former study utilized a full-length cDNA probe, including 3Ј-UTR se-quence, the higher molecular weight forms may represent 3Ј-UTR alternatively spliced variants. The smallest 1.7-kb transcript was not detected in the former study, perhaps because of the use of neonatal tissue in the current study versus adult tissues reported earlier, suggesting developmental regulation of this transcript. The difference in molecular weight of the two transcripts (⌬1.8 kb) correlates with the difference in length between the proximal and distal promoters. Thus, the distal P2-P4 promoter cluster initiated transcripts (which were not well resolved in our Northern analysis) likely correspond to the 3.5-kb transcript seen in Northern analysis with P1-initiated transcripts corresponding to the 1.7-kb product.
Once the rat ␣ 1a AR transcription start points were determined, alignment with published mouse and human ␣ 1a AR regulatory sequences was performed to gain insight into possible species-specific differences. Comparison with human ␣ 1a AR 5Ј-regulatory sequence reveals that although overall nucleotide identity is only 61%, several of the cis-elements such as MCAT, GATA, and AP1 within regions of increased identity are conserved (Fig. 9). Thus, although rat (and mouse) promoter sequence is quite divergent from human, the presence of similar putative cis-elements suggests that they can be transcribed under similar conditions. Indeed, unlike the murine promoter, the human ␣ 1a AR promoter is highly active in rat cardiomyocyte reporter gene assays (11-fold, Fig. 9). Interestingly, although rat and mouse promoters are highly similar with large stretches of identity (90%) and a number of conserved ciselements, including CRE, GATA, MCAT, and Sp1, the two promoters display large differences in promoter strength in the myocardium (32-fold versus 2.3-fold for full-length reporter constructs, Fig. 9). Further examination reveals that the most significant region of divergence falls in the segment containing the powerful rat basal enhancer, which contains putative binding sites for MCAT, GATA, and AP1 (light blue box, Fig. 9). This region shares little similarity to known sequences and may play a key regulatory role in the observed species-specific differences in ␣ 1a AR gene expression. The absence of this enhancer region in the human ␣ 1a AR gene appears to be compensated for with multiple other enhancers because rat and hu- man promoters result in very high overall expression, 32-fold and 11-fold over base line, respectively (compared with 2.3-fold for mouse).
The large difference in promoter strength between rat and mouse is consistent with markedly different regulation of ␣ 1a AR transcription between the two species in response to regulatory signals. For example, ␣ 1 AR agonist stimulation of rat cardiomyocytes results in up-regulation of endogenous rat ␣ 1a AR mRNA (13,14), whereas mouse ␣ 1a AR reporter constructs are virtually inert to ␣ 1 AR stimulation in these same cells (6). Also unlike the endogenous rat ␣ 1a AR gene, the mouse ␣ 1a AR promoter is unresponsive to stimulation by hypertrophic agonists such as endothelin-1, phorbol 12-myristate 13-acetate, and prostaglandin F2␣ (6). Additionally, rat and mouse ␣ 1a AR genes display very different responses to hypertrophy in vivo induced by aortic constriction pressure overload; such intervention results in up-regulation of ␣ 1a AR in rat models (14) but no change in mouse ␣ 1a AR levels (6,28). Finally, recent studies reveal that mouse ␣ 1 ARs also fail to activate phospholipase C␤, extracellular signal-regulated kinase, p38, or stimulate cardiomyocyte hypertrophy, indicating a generalized impairment of ␣ 1 AR signaling in murine myocardium (29). This is consistent with other studies demonstrating that several well characterized trophic effectors induce hypertrophy in rat cardiomyocytes (e.g. endothelin-1, phorbol 12-myristate 13-acetate, and prostaglandin F2␣) but have no effect on cultured mouse myocytes (30). Collectively, clear species-specific differences exist between mouse and rat cardiomyocytes in addition to differences in ␣ 1a AR promoter/gene structure which contribute to divergent regulation of ␣ 1a AR transcription in rat and mouse heart. Our current characterization of the rat ␣ 1a AR gene now makes it possible to address these differences mechanistically and should therefore provide insight into species-specific modulators of myocyte function.
Similar to human and mouse homologs, the rat ␣ 1a AR P1 and distal promoters lack both CAAT and TATA boxes. The distal promoters (P2-P4) occur within atypical initiation sites, with the predominant P2 transcript initiating within a GC box, P4 transcripts initiating at several sites within a 50-bp cluster, and the predominant P3 transcript initiating proximally to a 30-bp polypyrimidine tract. These tracts have been shown to regulate transcription at many levels including direct DNA binding (31,32), post-transcriptional processing of nascent transcripts (33), and alternate DNA structure/topology (34). The lack of typical regulatory or promoter elements for the distal promoters may explain the presence of multiple promoters in this region. Conversely, the P1 initiator conforms to a loose initiator sequence that is an identical match to a functional initiator element in the cdc25 gene (CCCAGCT) (35). From our serial deletion experiments, it is apparent that basal ␣ 1a AR promoter activity resides in most distal cloned 400-bp fragment, with the loss of this region abrogating 95% of basal transcriptional response. Additionally, P1 ␣ 1a AR transcription is modulated by multiple independent repressor (Ϫ681/Ϫ520, Ϫ520/Ϫ457 region) and activator (Ϫ1233/Ϫ950) elements in rat neonatal cardiomyocytes, containing a number of putative binding sites for transcription factors known to be relevant in heart, such as GATA, MCAT, and Sp1. The latter two sites have been shown to be important for basal expression of the human (5) and mouse (6) ␣ 1a AR genes, respectively, and may provide additional layers of regulatory control for the rat gene.
Although the functional relevance of multiple initiation sites is not yet clear, use of multiple promoters is common among several characterized ␣ 1 AR genes (rat ␣ 1b AR (8), human ␣ 1b AR (9), rat ␣ 1d AR (10)). For example, the rat ␣ 1b AR gene generates transcripts of 2.3, 2.7, and 3.3 kb in length via multiple pro-moters (8). Recent studies show that the proximal ␣ 1b AR promoter directs expression in a tissue-specific manner (36), with DNase footprinting experiments revealing that the proximal promoter p1f3 region binds a tissue-restricted factor in liver and hamster smooth muscle extracts, distinct from the predominant P2 promoter (36). In contrast, the recently identified murine ␣ 1a AR ortholog is transcribed from a single promoter located at Ϫ588 relative to the ATG (6). The human ␣ 1a AR ortholog, unlike either rat or mouse, is transcribed from a dominant promoter at Ϫ696 (relative to ATG), with several minor initiation sites clustered proximally (5). Given these data, a potential reason for multiple promoter utilization may be to aid in cell-specific regulation of a given gene, consistent with the well documented tissue-and species-specific distribution of ␣ 1 AR subtypes (4,20,37). Alternate transcripts might also be used to modulate ␣ 1a AR mRNA stability as has been reported recently for the rat ␤ 1 AR gene in response to agonist-independent and dependent down-regulation of ␤ 1 AR mRNA (38).
Myocardial hypertrophy occurs via several mechanisms. Well characterized among these mechanisms are pathways resulting from sympathetic (norepinephrine) stimulation (14) secondary to chronic hypoxia and/or ischemic insult (39). Recent studies aimed at elucidating mechanisms governing hypertrophy in response to these stimuli in rat heart reveal that ␣ 1a AR expression is modulated differentially by hypoxia (versus norepinephrine stimulation) in a chamber specific manner. This suggests that norepinephrine-stimulated hypertrophy proceeds via distinct pathway(s) from hypertrophy secondary to hypoxic stress (40) and is consistent with in vitro studies demonstrating hypoxia results in decreased ␣ 1a AR mRNA and receptor levels coupled with attenuation of ␣ 1a AR-mediated signaling and hypertrophy (15). Interestingly, this is in contrast to increased ␣ 1a AR mRNA levels with norepinephrine or phenylephrine stimulation in the same model (14). To begin to elucidate these distinct mechanisms, we analyzed transcriptional activity of the rat ␣ 1a AR gene in neonatal cardiomyocytes under normoxic and hypoxic conditions. Initially, we examined a potential role for the well characterized HIF-1, a heterodimeric, basic helix-loop-helix transcription factor expressed in response to cellular hypoxia which mediates multiple cellular and systemic homeostatic responses to hypoxia (41). In the rat ␣ 1a AR gene, although several regions are able to confer hypoxia-mediated regulation in transient reporter assays, the 5Ј-regulatory region is devoid of HIF-1 consensus sequences (42), suggesting that direct HIF-1 binding may not be the primary mechanism for altered (decreased) rat ␣ 1a AR expression with hypoxia. This is supported by in vitro gel shift analysis using anti HIF-1␣ antibody. Results showed that the Ϫ558/Ϫ542 region binds to nuclear factors under hypoxic conditions, and these shifts are not affected by the presence of anti-HIF. Functional analysis of mutant reporter constructs that alter this region in sequence reveals the presence of a novel 9-bp HRE (Ϫ555 CCCATGTCA Ϫ547) Further examination of these sequences reveals no obvious candidates known to confer hypoxia responsiveness, suggesting potentially novel pathways for hypoxia-mediated transcriptional control in rat ␣ 1a AR gene.
Although our findings do not preclude a potential indirect role of HIF-1, recent studies demonstrate that other factors modulate transcription during hypoxia, independent of direct HIF-1 binding. For example, it has been demonstrated that cAMP and subsequent CREB levels decrease under low oxygen exposure, contributing to hypoxia-elicited induction of epithelial tumor necrosis factor-␣ (43). Mechanistically, this may occur via phosphorylation of serine 133 by an unidentified kinase (44), resulting in CREB ubiquitination and proteosomal degradation (45). Given the presence of a cAMP response element at Ϫ756, CREB degradation could explain hypoxia-mediated repression of ␣ 1a AR transcriptional activity in the Ϫ950/ Ϫ681 region (i.e. loss of an activator). Indeed, previous studies in our laboratory have shown that the human ␣ 1a AR gene is induced by increased cAMP levels (5). Although direct binding of proteins to these regions was not seen in our gel shift analysis, it is possible that factors associating with the ␣ 1a AR HRE aid in nucleating complexes at these other sites. Such is the case with the yeast homothallic switching endonuclease gene promoter where transient Swi5 binding (Ͻ5 min) initiates a cascade of sequential binding of Swi/Snf and SAGA complexes at distal sites (46). This in turn recruits swi4/swi6-dependent cell-cycle box (SBF) to the promoter, which then initiates homothallic transcription in early G 1 phase (46). If a similar process regulates the ␣ 1a AR proximal promoter during hypoxia, no binding at the more proximal sites would be seen by gel shift analysis. Alternatively, another explanation is that deletion of more proximal sites alters helical phasing of factor binding within the Ϫ558/Ϫ520 region to the promoter, resulting in a decrease in transcriptional activity. Indeed, phasing, or location of protein binding on a face of the DNA helix, has been shown to be important in the collagenase-3 promoter where insertion of nucleotides to disrupt helical phasing between the AP-1 and runt domain sites results in abrogation of AP-1/runt domain binding and decreased collagenase-3 promoter activity (47). Studies are under way in our laboratory to determine which of these potential processes may be at work regulating ␣ 1a AR transcription with hypoxia.
In summary, we present the first cloning and characterization of the rat ␣ 1a AR gene, identifying a number of cardiac enriched putative cis-element binding sites that can direct robust expression from proximal and distal promoters. Heterogeneity of ␣ 1a AR mRNAs in rat heart appears to be the result of transcription initiation from multiple promoters. We have correlated functional activity of a novel 9-bp HRE between P1 and P2 to HIF-independent hypoxia-induced binding activity to this region. Loss of this region attenuates ␣ 1a AR responsiveness to transcription, and the factor(s) associated with this region may help nucleate other factors such as CREB to the ␣ 1a AR promoter. These novel findings should facilitate studies designed to elucidate ␣ 1 AR-mediated mechanisms involved in distinct myocardial pathologies as well as aid in elucidation of differential ␣ 1a AR regulation across species.