Regulation of Na,K-ATPase β1 Subunit Gene Transcription by Low External Potassium in Cardiac Myocytes

Expression of Na,K-ATPase activity is up-regulated in cells incubated for extended intervals in the presence of low external K+. Our previous data showed that exposure of cardiac myocytes to low K+ increased the steady-state abundance of Na,K-ATPase β1 subunit mRNA. In the present study we determined that incubation of primary cultures of neonatal rat cardiac myocytes with low K+ augmented Na,K-ATPase β1 gene expression at a transcriptional level and that this effect required extracellular Ca2+. The stimulatory effect of low K+ on Na,K-ATPase β1 gene transcription was not dependent on increased contractile activity of cardiac myocytes. Na,K-ATPase β1 5′-flanking region deletion plasmids used in transient transfection analysis demonstrated that the region between nucleotides −62 to −42 of the β1 promoter contained a low K+response element. Site-directed mutagenesis of a potential GC box core motif GCG in the −58/−56 region of the β1 promoter decreased basal and low K+-mediated transcription. Mutation of the core sequence of a putative GC box element located between nucleotides −101 and −99 further decreased the low K+ effect on β1 gene transcription. Electrophoretic mobility shift assays using oligonucleotides spanning the proximal and distal GC box elements of the β1 promoter showed enhanced binding of two complexes in response to low K+. The inclusion of a consensus GC box sequence as a competitor in gel shift analysis reduced factor binding to the low K+ response elements. Antibodies to transcription factors Sp1 and Sp3 interacted with components of both DNA-binding complexes and binding of nuclear factors was abolished in gel shift studies using GC box mutants. Together these data indicate that enhanced binding of Sp1 and Sp3 to two GC box elements in the rat Na,K-ATPase β1 subunit gene promoter mediates β1 gene transcription up-regulation in neonatal rat cardiac myocytes exposed to low external K+.

Na,K-ATPase 1 is an intrinsic plasma membrane enzyme that plays an essential role in animal cell physiology (1). By its continuous function coupled to the hydrolysis of ATP, Na,K-ATPase is directly responsible for maintaining transmembrane gradients of Na ϩ and K ϩ , resting membrane potentials, control of cell volume, and a significant portion of basal energy utilization. By virtue of its predominant role in the regulation of intracellular Na ϩ , Na,K-ATPase is also of vital importance in the control of the intracellular concentrations of calcium and hydrogen ions whose transport are coupled to Na ϩ . Na,K-ATPase consists of two noncovalently linked, dissimilar subunits, ␣ and ␤, present in equimolar amounts (2). The larger ␣ subunit (112 kDa) regulates catalytic activity and contains sites for cardiac glycoside binding and phosphorylation (2). The smaller ␤ subunit (32 kDa) is extensively glycosylated and facilitates the correct insertion of the ␣ subunit in the plasma membrane (3). While multiple isoforms of Na,K-ATPase ␣ (␣1, ␣2, and ␣3) and ␤ (␤1, ␤2, and ␤3) subunits are expressed in a tissue-specific fashion (4 -8), ␣1 and ␤1 subunits are constitutively expressed in the majority of tissues (1).
Prolonged exposure of a variety of cells to conditions which are inhibitory to Na,K-ATPase function, such as digitalis derivatives or low extracellular K ϩ , yields an initial increase in the intracellular concentration of Na ϩ and a subsequent upregulation of Na,K-ATPase activity and subunit content (9 -12). Similar increases in Na,K-ATPase expression have been noted in vivo with administration of digoxin to experimental animals (13)(14)(15) and humans (16,17). A pretranslational mechanism has been implicated in the up-regulation of Na,K-ATPase activity based on the finding of either low K ϩ -or ouabain-mediated augmentation of Na,K-ATPase subunit mRNA contents in established cell lines such as Madin-Darby canine kidney and ARL-15 (18,19) and in primary cell cultures derived from adult rat kidney (20) and neonatal rat ventricle (21). Our previous transient transfection studies demonstrated that the region between Ϫ102 to ϩ151 base pairs of the rat Na,K-ATPase ␤1 subunit gene is necessary for up-regulation of ␤1 mRNA content in response to incubation of primary cultures of neonatal rat cardiac myocytes in low extracellular K ϩ (21). It is important to note, however, that neither the DNA sequence element(s) nor the trans-acting factor(s) responsible for control of Na,K-ATPase subunit gene expression in response to either low K ϩ or ouabain have been defined.
In the present studies, we have further delineated the molecular mechanism underlying regulation of Na,K-ATPase ␤1 gene expression in response to low extracellular K ϩ in primary cultures of neonatal rat cardiac myocytes. Our results indicate that low K ϩ mediates an increase in ␤1 subunit gene transcription that is dependent on extracellular Ca 2ϩ . Transient transfection studies with 5Ј deletion and mutant constructs revealed that two GC box elements in the proximal ␤1 gene promoter between nucleotides Ϫ102 and Ϫ42 are necessary for transcriptional up-regulation mediated by low K ϩ . The results of electrophoretic mobility shift and transient transfection assays show that increased binding of Sp1 and Sp3 transcription factors to these GC box sequences are required to elicit the low K ϩ response.

EXPERIMENTAL PROCEDURES
Materials-Neonatal rats (1-day-old) of the Harlan Sprague-Dawley strain were purchased from Charles River Laboratories (Wilmington, MA). Dulbecco's modified Eagle's medium, K ϩ -free and K ϩ -and Ca 2ϩfree Williams' E medium, serum, Lipofectin, oligonucleotides, and nicktranslation kits were obtained from Life Technologies, Inc. (Grand Island, NY). Luciferase assay systems and Taq polymerase were purchased from Promega (Madison, WI). Protein assay kits were from Bio-Rad. [␣-32 P]dCTP (3000 Ci/mmol) was from NEN Life Science Products, Inc. (Boston, MA) and BA-85 nitrocellulose filters from Schleicher & Schuell (Keene, NH). Site-directed mutagenesis kits were from Stratagene (La Jolla, CA). Ultraspec RNA isolation systems were from Biotecx Laboratories (Houston, TX). Sp1, Sp3, and USF-1 antibodies were obtained from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). All chemicals were purchased from Sigma and were analytical grade or better.
Plasmids-Construction of luciferase plasmids containing the Ϫ102 to ϩ151 and Ϫ41 to ϩ151 base pair regions of the rat Na,K-ATPase ␤1 gene was described previously (21,22). The luciferase reporter plasmids containing the promoter regions of the chicken myosin light chain-2 (MLC-2) (Ϫ222 to ϩ158) and rat angiotensinogen (Ϫ688 to ϩ39) genes were provided by M. A. Q. Siddiqui and A. Brasier, respectively. The generation of a luciferase plasmid harboring the Ϫ157 to Ϫ38 base pair region of the ␤1 gene upstream of the minimal promoter of the mouse mammary tumor virus (MMTV) has been reported (22). Luciferase constructs that contain ␤1 promoter sequences from Ϫ84 to ϩ151 and Ϫ62 to ϩ151 base pairs were synthesized using polymerase chain reaction (PCR) amplification. Oligonucleotides from Ϫ84 to ϩ151 and Ϫ62 to ϩ151 base pairs were generated by PCR using the Ϫ102/ϩ151 ␤1 region as a template. The reactions were performed in a total volume of 100 l containing 1 g of template, 0.2 mM dNTPs, 2 M of each primer (5Ј primers, GCGGATCCGATTGGCCTGCGGT and GCGGATC-CTAGGCGGAGCTAC; 3Ј primer, GCAAGCTTGCATGCAGG) 2.5 mM MgCl 2 , 50 mM KϩCl, 10 mM Tris-HCl (pH 9.0), 0.1% Triton X-100, and 2.5 units of Taq polymerase. 25 cycles were run, each cycle included 94°C for 15 s (denaturation), 70°C for 15 s (annealing), and 72°C for 1 min and 15 s (elongation). The PCR fragments were digested with BamHI and ligated into the BamHI site of the luciferase vector pXP1. The PCR to synthesize the Ϫ55 to ϩ151 oligonucleotide was performed in a total volume of 50 l containing 30 ng of template, 0.1 mM dNTPs, 0.1 nM of each primer (5Ј primer, AAAGAATCCGAGCTACGGATG-GTGGAGGC; 3Ј primer, CTTGAATCCCCTGCTGCTTCAAG), 2 mM MgCl 2 , 10 mM (NH 4 ) 2 SO 4 , 10 mM KCl, 20 mM Tris-HCl (pH 8.8), 0.1% Triton X-100, and 2.5 units of Pfu polymerase (Stratagene). The run included 28 cycles, each cycle included: denaturation for 1 min and 30 s (96°C), annealing for 40 s (60°C) and elongation for 1 min and 30 s (68°C). The PCR fragment was ligated into pXP1. Each 5Ј deletion clone was sequenced to confirm the appropriate sequence.
Site-directed Mutants-A single GC box mutant that changed GCG to TTT in the Ϫ58/Ϫ56 region of the ␤1 promoter was generated using the QuikChange site-directed mutagenesis kit according to the manufacturer's protocol (Stratagene). The luciferase plasmid containing the ␤1 gene Ϫ102 to ϩ151 region was used as a template with two synthetic oligonucleotides containing the GC box mutation (forward primer, CG-GTGCCGCCGGTAGTTTGAGCTACGGATGGTGGAG; reverse primer, CTCCACCATCCGTAGCTCAAACTACCGGCGGCACCG). A double mutant that changed the Ϫ58/Ϫ56 and Ϫ67/Ϫ65 sequences from GCG and CGC to AGA and TCT, respectively, was produced using the Chameleon site-directed mutagenesis kit (Stratagene) and mutant primers (forward primer, CGATTGGCCTGCGGTGCTCTCGGTAGAGAGAGCTA-CGGATGG; reverse primer, CCATCCGTAGCTCTCTCTACCGAGAGC-ACCGCAGGCCAATCG). PCR amplification was carried as detailed in the manufacturer's protocol with 10 ng of luciferase plasmid containing the ␤1 gene Ϫ102 to ϩ151 region as a template and 125 ng of each primer. This double mutant was used as a template to generate a triple GC box mutant with the additional change of nucleotides at Ϫ101/Ϫ99 from CGC to TCT using the QuikChange site-directed mutagenesis kit and mutant primers (forward primer, GCTCAGATCTCGAGGCCCTC-TCTTCTCGGCACCGGC; reverse primer, GCCGGTGCCGAGAAGAG-AGGGCCTCGAGATCTGAGC). All site-directed mutants were confirmed by DNA sequence analysis.
Cell Culture and Transfection-Neonatal rat cardiac myocytes were prepared and incubated as described previously (21). Briefly, ventricles from 1 to 3-day-old Harlan Sprague-Dawley rats were minced and subjected to consecutive digestion with trypsin and DNase I. Cells were preplated for 30 min at 37°C and myocytes (2826/mm 2 ) remaining in suspension were added to plates coated with 1% gelatin. Cardiac myocytes were incubated overnight in Dulbecco's modified Eagle's medium containing 10% calf serum, 0.1 mM bromodeoxyuridine, and antibiotics at 37°C in humidified air with 7% CO 2 . In the majority of experiments, cells were transfected by an overnight incubation at 3% CO 2 with calcium phosphate-DNA coprecipitates. To minimize the potential effect of increased intracellular Ca 2ϩ on transfection analysis, in one set of experiments cardiac myocytes were transfected overnight with Lipofectin (Life Technologies, Inc.) according to the manufacturer's protocol. In experiments designed to evaluate the effect of contraction on transfected gene expression, cells were washed with 1 ϫ phosphate-buffered saline and then incubated for 24 h in serum-free, K ϩ -free Willliams' E medium supplemented to a final concentration of 5.4 mM K ϩ with either 54 mM K ϩ , 10 M verapamil, or 7.5 mM 2,3-butanedione monoxime (BDM). In all other transfection experiments, cardiac myocytes were preincubated for 24 h in serum-free, K ϩ -free Willliams' E medium supplemented to a final concentration of 5.4 mM K ϩ and then exposed to reduced concentrations of K ϩ for 1 day. Cells were lysed and luciferase and ␤-galactosidase activities were measured in cell extracts as described previously (21). Protein concentration was quantitated using a Bio-Rad assay system. It was determined that the expression of a co-transfected CMV/␤-galactosidase vector was significantly regulated by exposure to low K ϩ , high K ϩ , verapamil or BDM. In three experiments in which four plates were transfected with the Ϫ102/ϩ151 ␤1/ luciferase plasmid, we found that the variation of luciferase activity expressed relative to total protein was 7 Ϯ 0.7% (p Ͻ 0.05). Luciferase activities were therefore normalized to total cellular protein.
Northern Blot Hybridization-Neonatal rat heart cells were incubated in Dulbecco's modified Eagle's medium containing 10% calf serum for 2 days and then preincubated for 24 h in serum-free, K ϩ -free Willliams' E medium supplemented to a final concentration of 5.4 mM K ϩ . Cells were washed with 1 ϫ phosphate-buffered saline and incubated in K ϩ -free Williams' E medium supplemented with either 5.4 or 0.3 mM K ϩ for 6 h. In parallel, cardiac myocytes were incubated with 0.1 mM EGTA in the presence and absence of 2.1 mM Ca 2ϩ for 6 h and the effect of low K ϩ was assessed. Cells were lysed with 5 ml of Ultraspec RNA isolation reagent (Biotecx Laboratories, Inc.) containing guanidinium thiocyanate and phenol and Northern blot analysis was conducted as previously detailed (21). Briefly, total RNA (10 g) was resolved on 1% agarose gels containing 5% formaldehyde and RNA was transferred to nitrocellulose membranes. Blots were prehybridized overnight at 42°C in 5 ϫ SSC (1 ϫ SSC ϭ 0.15 M NaCl, 0.015 M sodium citrate, pH 7.0), 5 ϫ Denhardts, 50% formamide, 1 mM EDTA, 10 mM NaH 2 PO 4 (pH 6.9), 0.5% SDS, and 0.1 mg/ml sonicated, sperm DNA. A Na,K-ATPase ␤1 cDNA was labeled with 32 P by nick-translation and blots were hybridized in a solution containing 5 ϫ SSC, 1 ϫ Denhardts, 50% formamide, 1 mM EDTA, 10 mM NaH 2 PO 4 (pH 6.9), 0.1% SDS, and 0.1 mg/ml sperm DNA. After a 3-day incubation at 42°C, membranes were washed with 0.1 ϫ SSC, 0.1% SDS twice at 55°C for 15 min. Na,K-ATPase ␤1 subunit mRNA was detected and quantitated by Phos-phorImaging analysis.
Preparation of Nuclear Extract-The method used for preparing nuclear extracts was a modified Dignam protocol (23). Cells (25 ϫ 10 6 ) were washed twice with ice-cold 1 ϫ phosphate-buffered saline, scraped into 1.5 ml of ice-cold 1 ϫ phosphate-buffered saline in a 1.5-ml Eppendorf tube and centrifuged at 2,000 rpm for 10 s at room temperature. To rupture cell membranes, pellets were resuspended in 5 volumes of an ice-cold buffer containing 10 mM HEPES (pH 7.9), 10 mM KCl, 1.5 mM MgCl 2 , 0.5 mM dithiothreitol, 0.5 mM phenylmethylsulfonyl fluoride, and 0.5 g/ml leupeptin, incubated on ice for 10 min and vortexed vigorously for 20 s. The homogenate was subjected to centrifugation at 2,000 rpm for 10 s. Nuclear pellets were resuspended in an equal volume of a buffer comprised of 20 mM HEPES (pH 7.9), 420 mM NaCl, 1.5 mM MgCl 2 , 0.2 mM EDTA, 25% glycerol, 0.5 mM phenylmethylsulfonyl fluoride, 0.5 mM dithiothreitol, and 0.5 g/ml leupeptin and incubated on ice for 20 min. After centrifugation at 14,000 rpm for 2 min at 4°C, the supernatant containing nuclear proteins was aliquoted and stored at Ϫ70°C. Protein concentration was determined using a Bio-Rad kit.
Electrophoretic Mobility Shift Assay-Double-stranded oligonucleotides spanning three potential GC box elements in the rat Na,K-ATPase ␤1 gene promoter were commercially synthesized. In addition, corresponding oligonucleotides were prepared with the core GCG sequences mutated to either TTT or AGA. Wild type and mutant oligonucleotides were labeled using the Klenow fragment in the presence of [␣-32 P]dCTP and used as probes. Nuclear extract isolated from cardiac myocytes cultured in either 5.4 or 0.3 mM K ϩ for 1 day was incubated in a 15-l reaction with 1 ng of 32 P-labeled oligonucleotide (2 to 5 ϫ 10 5 cpm) in 10 mM HEPES (pH 7.9), 75 mM KCl, 1 mM EDTA, 1 mM dithiothreitol, 10% glycerol, and 1.5 g of poly(dI-dC) for 15 min at room temperature. For oligonucleotide competition experiments, a 100-fold excess of unlabeled oligonucleotide competitors were preincubated with nuclear extracts for 15 min at 4°C prior to the addition of labeled probe and incubation for 15 min at room temperature. For supershift experiments, antibodies were preincubated with nuclear extracts for 15 min at 4°C and then incubated with probe at 4°C for an additional 15 min. Reaction mixtures were applied to a 6% polyacrylamide gel and DNA-protein complexes were separated by electrophoresis under low ionic strength conditions (0.5 ϫ Tris borate-EDTA buffer) at 15 mA at 4°C. The gel was dried under vacuum and analyzed by PhosphorImaging.
Statistical Analysis-Student's t tests and ANOVA were utilized to determine the significance of results.

Low K ϩ Induces a Selective, Promoter-dependent Increase in
Na,K-ATPase ␤1 Gene Transcription-Our previous study (21) showed that incubation of primary cultures of neonatal rat cardiac myocytes with low extracellular K ϩ increased the steady-state level of Na,K-ATPase ␤1 mRNA and enhanced luciferase activity driven by the Ϫ102 to ϩ151-base pair region of the rat ␤1 gene. To further delineate the molecular mechanism which underlies the low K ϩ effect on Na,K-ATPase ␤1 gene expression, neonatal rat cardiac myocytes were transiently transfected with luciferase reporter plasmids containing portions of the rat ␤1 promoter between Ϫ102 to ϩ151 or Ϫ41 to ϩ151 base pairs. Exposure of cells to 0.3 mM K ϩ for 1 day resulted in a 2.6 Ϯ 0.3-fold increase in luciferase activity by the Ϫ102 to ϩ151-base pair region, whereas the Ϫ41 to ϩ151 portion of the ␤1 promoter was unresponsive to low K ϩ (Fig.  1A). These data indicate that the low K ϩ response is transcriptional and is mediated by a DNA element between nucleotides Ϫ102 to Ϫ42 of the rat Na,K-ATPase ␤1 gene. To begin to assess the selectivity of the low K ϩ response, we examined the effect of low K ϩ on MLC-2 and angiotensinogen transcription in transient transfection experiments (Fig. 1B). In cells treated with 0.3 mM K ϩ for 1 day, luciferase activity directed by the rat angiotensinogen gene promoter was unaffected by low K ϩ . By contrast, luciferase expression was augmented 4.8 Ϯ 0.5-fold by low K ϩ in cells transfected with a reporter plasmid containing the proximal promoter region of the chicken MLC-2 gene. To examine whether the Ϫ102 to Ϫ42 region of the ␤1 gene functions as a promoter-independent enhancer in mediating the low K ϩ effect, we subcloned the Ϫ157 to Ϫ38-base pair region of the ␤1 gene into a luciferase vector driven by the proximal promoter of MMTV and evaluated its response to low K ϩ (Fig. 1C). Interestingly, low K ϩ treatment had no stimulatory effect on this FIG. 1. Low K ؉ induces a selective, promoter-dependent stimulation of rat Na,K-ATPase ␤1 gene transcription in neonatal rat cardiac myocytes. Panel A, primary cultures of neonatal rat cardiac myocytes were prepared and transfected by calcium phosphate coprecipitation with 1 g of luciferase plasmids containing either the Ϫ102 to ϩ151 or Ϫ41 to ϩ151-base pair region of the rat Na,K-ATPase ␤1 gene as detailed under "Experimental Procedures." Cells were incubated in serum-free Williams' E medium in either 5.4 or 0.3 mM K ϩ for 1 day. Cardiac myocytes were lysed and cell extracts assayed for luciferase activity. Panel B, cardiac myocytes were transfected with 1 g of luciferase plasmids harboring portions of the 5Ј ends of the rat angiotensinogen and chicken MLC-2 genes. Transfected cells were incubated in the presence of either 5.4 or 0.3 mM K ϩ for 1 day. Values are presented as the ratio of luciferase activity in 0.3 mM K ϩ relative to 5.4 mM K ϩ . Panel C, a luciferase construct containing the Ϫ157 to Ϫ38-base pair region of the ␤1 subunit gene upstream of the MMTV promoter was transfected into neonatal rat cardiac myocytes. Luciferase plasmids containing either the MMTV promoter or the Ϫ102 to ϩ151-base pair region of the ␤1 subunit gene were introduced into replicate plates of cells. Cells were cultured in serum-free medium containing either 5.4 or 0.3 mM K ϩ for 1 day and luciferase activity was determined as described above. For panels A-C, data from three to four independent experiments were pooled (n ϭ 6 to 8). *, p Ͻ 0.05. heterologous promoter suggesting that the low K ϩ response element requires the native ␤1 gene promoter to function.
Partial Inhibition of Na,K-ATPase Activity and Extracellular Ca 2ϩ Are Required for Low K ϩ Stimulation of Na,K-ATPase ␤1 Gene Expression-A physiological reduction in serum K ϩ in the range of 2 to 3 mM has been shown to regulate Na,K-ATPase subunit gene expression in a tissue-dependent manner (24). By contrast, incubation of cardiac myocytes in a medium containing a K ϩ concentration of 1 mM or lower leads to Na,K-ATPase inhibition (10). To determine whether Na,K-ATPase ␤1 subunit gene transcription was regulated by either a physiological decrease in K ϩ concentration or a prolonged inhibition of Na,K-ATPase function, the effect of several concentrations of K ϩ on ␤1 gene activity was assessed in transient transfection experiments in primary cultures of neonatal rat cardiac myocytes (Fig. 2). Relative to control cells incubated with 5.4 mM K ϩ , treatment with 2 mM K ϩ had no effect on luciferase activity of the Ϫ102 to ϩ151 base pair construct while exposure to 1, 0.6, or 0.3 mM K ϩ for 1 day resulted in stimulation of chimeric gene expression. Thus, a physiological concentration of K ϩ , 2 mM, did not stimulate Na,K-ATPase ␤1 expression, suggesting that partial inhibition of Na,K-ATPase activity due to a limiting concentration of K ϩ was a necessary first step in the signal transduction pathway which ultimately up-regulates ␤1 gene transcription.
Partial inhibition of Na,K-ATPase function by low external K ϩ yields an increase in the intracellular concentration of Na ϩ and a subsequent elevation in cytoplasmic Ca 2ϩ as a consequence of reduced Na ϩ /Ca 2ϩ exchange activity (10). Given the central role that Ca 2ϩ plays in a myriad of signal transduction systems, we evaluated the effect of the absence of extracellular Ca 2ϩ on the low K ϩ -mediated enhancement of Na,K-ATPase ␤1 gene expression. We first incubated neonatal rat cardiac myocytes in the absence and presence of Ca 2ϩ and assessed the effect of low K ϩ on ␤1 mRNA content (Fig. 3, A and B). In the presence of a normal concentration of Ca 2ϩ (2.1 mM), incubation of cells for 6 h with 0.3 mM K ϩ yielded a 1.3 Ϯ 0.1-fold increase in Na,K-ATPase ␤1 mRNA abundance. The low K ϩ effect was abrogated, however, when cardiac myocytes were cultured in Ca 2ϩ -free medium containing 0.1 mM EGTA and then challenged with 0.3 mM K ϩ for 6 h. Thus, low K ϩ stimulation of ␤1 mRNA levels requires extracellular Ca 2ϩ . It is also important to note the withdrawal of Ca 2ϩ was associated with a significant decrease in the content of ␤1 mRNA, suggesting that Ca 2ϩ by itself is capable of regulating Na,K-ATPase ␤1 gene expression. To further investigate the role of Ca 2ϩ in the low K ϩ effect on ␤1 gene transcription, we compared the effect of low K ϩ in the absence and presence of extracellular Ca 2ϩ on the expression of the Ϫ102 to ϩ151 ␤1 gene construct (Fig. 3C). Treatment of cardiac myocytes with 0.3 mM K ϩ for 6 h had no effect on luciferase activity, either in the presence or absence of extracellular Ca 2ϩ , when cells were transfected with a standard calcium phosphate method. An extracellular Ca 2ϩ -dependent 1.5 Ϯ 0.1-fold increase in luciferase activity was produced in response to a 6-h exposure to low K ϩ , however, when cells were transfected with Lipofectin. This result is consistent with the finding that the low K ϩ effect on ␤1 mRNA requires extracellular Ca 2ϩ . The reason for the discrepancy in the transient transfection results with calcium phosphate versus Lipofectin transfection is not known. We speculate that the influx of Ca 2ϩ that occurs with the calcium phosphate transfection procedure and the early time point of analysis may be involved, since we have demonstrated in transient transfection experiments utilizing either calcium phosphate coprecipitation or Lipofectin that exposure to 0.3 mM K ϩ for 24 h augments ␤1 chimeric gene expression to the same extent (data not shown).
A 21-Base Pair Sequence of the Na,K-ATPase ␤1 Gene Promoter Is Necessary for Low K ϩ -mediated Stimulation of ␤1 Gene Transcription-To initiate an investigation of the DNA sequence motif involved in low K ϩ -induced activation of Na,K-ATPase ␤1 gene expression, we carried out a computer-assisted examination of the Ϫ102 to Ϫ42 base pair low K ϩ response region for the presence of potential transcription factor binding elements (Fig. 4A). Putative CAAT and AP-2 motifs were identified at positions Ϫ84 and Ϫ78, respectively, relative to the transcription start site of the ␤1 gene. In addition, a potential GC box and E box were located at positions Ϫ61 and Ϫ49, respectively. E box elements are binding sites for members of the basic helix-loop-helix leucine zipper family of transcription factors (25). Moreover, the binding of a basic helix-loop-helix leucine zipper protein to an E box motif has been shown to be necessary for contractile-mediated up-regulation of the rat ␣-myosin heavy chain promoter (26). Since low K ϩ treatment of neonatal rat cardiac myocytes leads to enhanced contractile activity and the low K ϩ response region contains a potential E box element, we evaluated the effect of contraction on Na,K-ATPase ␤1 gene transcription in transient transfection studies (Fig. 4B). Cultures of spontaneously contracting neonatal rat cardiac myocytes were transfected with the Ϫ102 to ϩ151 ␤1 promoter construct and treated for 24 h with 54 mM K ϩ , 10 M verapamil, or 7.5 mM BDM in serum-free medium. All three interventions inhibit contractile activity of cultured cardiac myocytes; however, they exert distinct effects on intracellular Ca 2ϩ (27). High K ϩ causes membrane depolarization and elevates intracellular Ca 2ϩ whereas verapamil and BDM decrease the intracellular concentration of Ca 2ϩ and reduces Ca 2ϩ transients, respectively. Relative to luciferase expression in spontaneously contracting control cells incubated with 5.4 mM K ϩ , luciferase activity in cardiac myocytes exposed to 54 mM K ϩ increased 3.7 Ϯ 0.4-fold. By contrast, verapamil and BDM treatment resulted in a significant decrease in Na,K-ATPase ␤1 promoter activity. Since all three agents effectively suppressed contract-  4. Sequence of the ؊102 to ؊42-base pair region of the rat Na,K-ATPase ␤1 gene and the effect of contraction on ␤1 gene transcription. Panel A, potential transcription factor binding sites in the low K ϩ response region are underlined. R refers to the reverse complement. Panel B, neonatal rat cardiac myocytes were transfected with a luciferase plasmid containing the Ϫ102/ϩ151 ␤1 gene region and incubated in serum-free medium as described under "Experimental Procedures." Cells were exposed to either 10 M verapamil or 7.5 mM BDM in the presence of 5.4 mM K ϩ or incubated with 54 mM K ϩ for 1 day. Luciferase activities are expressed relative to the value derived from cells exposed to 5.4 mM K ϩ . Data from four independent experiments done in duplicate were pooled (n ϭ 8). *, p Ͻ 0.05. ile activity, it appears that low K ϩ up-regulation of Na,K-ATPase ␤1 gene transcription is contraction-independent.
To determine the precise DNA motif within the Ϫ102 to Ϫ42-base pair region of the ␤1 promoter which mediates low K ϩ up-regulation of Na,K-ATPase ␤1 subunit gene transcription, additional 5Ј-flanking deletion constructs were prepared. We first quantitated basal expression of ␤1 promoter deletion plasmids in serum-free medium containing 5.4 mM K ϩ (Fig. 5A). The ␤1 Ϫ84 and Ϫ62 deletion constructs exhibited 60 and 42%, respectively, of the promoter activity of the Ϫ102 plasmid. Further deletion of ␤1 5Ј-flanking sequence to Ϫ41 resulted in 9% luciferase expression relative to the Ϫ102 to ϩ151 base pair construct. In cells exposed to 0.3 mM K ϩ for 1 day, ␤1 promoter activity was stimulated 1.8 -1.9-fold in ␤1 deletion constructs containing 5Ј-flanking sequences from Ϫ102 to Ϫ62, whereas, consistent with data presented in Fig. 1, low K ϩ treatment had no affect on the Ϫ41 to ϩ151 ␤1 construct (Fig. 5B). These data demonstrate that a low K ϩ response element resides in the Ϫ62 to Ϫ42-base pair region of the ␤1 gene. To examine whether the putative GC box located between nucleotides Ϫ61 and Ϫ52 of the ␤1 promoter is necessary for the low K ϩ effect, an additional deletion mutant containing the Ϫ55 to ϩ151-base pair region of the Na,K-ATPase ␤1 gene was prepared and used in transient transfection studies (Fig. 5C). Basal expression of the Ϫ55 plasmid in the presence of 5.4 mM K ϩ was reduced 77% relative to the Ϫ62 base pair reporter construct, and the low K ϩ -mediated stimulation of luciferase expression was abolished in the Ϫ55 ␤1 deletion mutant. Taken together, these results indicate that a functional GC box in the Ϫ61 to Ϫ52 region of the Na,K-ATPase ␤1 promoter plays a major role in basal expression and mediates stimulation of ␤1 gene transcription in response to low K ϩ .
FIG. 5. Basal and low K ؉ -induced transcription of 5 deletion constructs of the ␤1 promoter. Panel A, neonatal rat cardiac myocytes were incubated and transfected as described in the legend to Fig. 1 with 1 g of luciferase plasmids harboring 5Ј deletions of the 5Ј-flanking region of the ␤1 gene. Luciferase activities are expressed relative to the value determined for the Ϫ102 to ϩ151 plasmid which is set to 1.0. Panel B, cells were transfected as in A and incubated in the presence of either 5.4 or 0.3 mM K ϩ for 1 day. Luciferase activities in cells incubated in 0.3 mM K ϩ are expressed relative to the value derived from cells exposed to 5.4 mM K ϩ . Panel C, cardiac myocytes were incubated and transfected as in A with 5 g of luciferase plasmids containing either the Ϫ62 to ϩ151 or Ϫ55 to ϩ151-base pair region of the ␤1 gene. Cells were exposed to either 5.4 or 0.3 mM K ϩ for 1 day and luciferase activity was assayed. Luciferase activities are expressed relative to the value obtained for the Ϫ62 to ϩ151 plasmid in cells treated with 5.4 mM K ϩ . Panel D, cells were transfected as in C and incubated in either 5.4 or 54 mM K ϩ for 1 day. Luciferase activities are expressed relative to the value derived from cells exposed to 5.4 mM K ϩ . Data shown in Panels A and B were derived from seven to eight independent experiments done in duplicate (n ϭ 14 -16). Data depicted in Panels C and D were from three independent experiments done in duplicate (n ϭ 6). *, p Ͻ 0.05.
To assess whether an increase in the intracellular concentration of Ca 2ϩ by itself is the underlying mechanism of the effect of low K ϩ , neonatal rat cardiac myocytes were transfected with the Ϫ62 and Ϫ55 ␤1 deletion constructs and cells were incubated with either 5.4 or 54 mM K ϩ for 1 day (Fig. 5D). Exposure to 54 mM K ϩ and the resultant increase in intracellular Ca 2ϩ stimulated luciferase expression from both deletion constructs; thereby suggesting that elevated intracellular Ca 2ϩ is not solely responsible for up-regulation of Na,K-ATPase ␤1 gene transcription in response to low K ϩ . This result also indicated that Ca 2ϩ responsiveness is mediated by the Ϫ55 to ϩ151-base pair region of the ␤1 gene.
Sp1 and Sp3 Bind to a Na,K-ATPase ␤1 Gene Low K ϩ Response Element-To investigate protein binding to the region of the ␤1 promoter which is necessary for low K ϩ -induced augmentation of ␤1 gene transcription, nuclear extracts were prepared from neonatal rat cardiac myocytes incubated in either 5.4 or 0.3 mM K ϩ for 1 day. Electrophoretic mobility shift assays were conducted with nuclear extracts and an oligonucleotide corresponding to the Ϫ67 to Ϫ35-base pair region of the Na,K-ATPase ␤1 gene. In both control and low K ϩ -treated cells two classes of protein-DNA complexes were detected and binding was enhanced by low K ϩ (Fig. 6A). Quantitation of repeated electrophoretic mobility shift assays revealed that low K ϩ yielded a 2.1 Ϯ 0.4-and 3.0 Ϯ 0.6-fold increase in the intensity of lower and upper protein-DNA complexes, respectively (Fig. 6B). Interestingly, incubation of neonatal rat cardiac myocytes with low K ϩ for 1 day was associated with a 26 Ϯ 7% increase in the recovery of total nuclear protein (n ϭ 10, p Ͻ 0.05).
To identify proteins that bind to the low K ϩ response element of the ␤1 promoter, oligonucleotide competition analysis was conducted (Fig. 7A). Both lower and upper protein-DNA complexes were significantly reduced when gel shift assays contained either unlabeled Ϫ67 to Ϫ35 ␤1 gene oligonucleotide (lanes 3 and 4) or oligonucleotide Sp1 containing the consensus GC box sequence (lanes 7 and 8). By contrast, the formation of the two complexes was not affected when an oligonucleotide containing an USF-1-binding site was used as a competitor (lanes 5 and 6). These results suggest that low K ϩ -induced protein binding occurs within the GC box of the low K ϩ response region of the ␤1 promoter.
To determine which member of the GC box-binding proteins was involved in the low K ϩ response, electrophoretic mobility shift assays were performed with 1 g of nuclear extracts and 1 g of antibodies for Sp1, Sp3, and USF-1 (Fig. 7, B and C). The lower protein-DNA complex was abolished in the presence of anti-Sp3 antibody (Fig. 7B, lanes 5 and 6). By contrast, lower band formation was unaffected by either anti-Sp1 antibody (lanes 3 and 4), anti-USF-1 antibody (lanes 7 and 8), or IgG (lanes 9 and 10). Quantitation of the effect of anti-Sp1 antibody on the upper protein-DNA complex revealed a 30% decrease in DNA binding in nuclear extracts derived from low K ϩ -treated cardiac myocytes (Fig. 7C). Inclusion of Sp3-specific antibody reduced the upper protein-DNA complex 42 and 60% in extracts from control and low K ϩ -treated cells, respectively (Fig. 7C). Neither anti-USF-1 antibody nor IgG affected upper band formation. Essentially identical results were obtained in electrophoretic mobility shift assays conducted with 1 g of nuclear extract and 4 g of anti-Sp1 and anti-Sp3 antibodies (data not shown). These data indicated that the lower protein-DNA complex was formed by the binding of Sp3 while the upper complex represented a combination of both Sp1 and Sp3. When the core GCG motif of the GC box was mutated to TTT (Fig. 7D, lanes 3 and 4), both lower and upper band formation were abolished, supporting the contention that binding of Sp1 and Sp3 are necessary to mediate the up-regulation of Na,K-ATPase ␤1 gene transcription in response to low K ϩ .
Role of Additional Putative GC Box Motifs in Regulation of ␤1 Gene Transcription by Low K ϩ -Within the Ϫ102 to Ϫ42base pair region of the promoter of the rat Na,K-ATPase ␤1 gene there exists multiple potential GC box elements (Fig. 8A). To assess the functional role of these putative GC box motifs in low K ϩ -mediated transcriptional up-regulation of the ␤1 gene, we mutated the core GCG sequence to generate the luciferase FIG. 6. Low K ؉ increases binding to the ؊67 to ؊35-base pair region of the ␤1 promoter. Panel A, exposure of cardiac myocytes to low K ϩ , nuclear extract preparation and electrophoretic mobility shift assay are described under "Experimental Procedures." Representative binding of 3 g of nuclear extracts derived from control cells (lanes 1 and 2) and cardiac myocytes exposed for 1 day to 0.3 mM K ϩ (lanes 3 and  4) to the labeled Ϫ67/Ϫ35 ␤1 oligonucleotide. Panel B, quantitation of the effect of low K ϩ on upper and lower band formation on the Ϫ67/Ϫ35 ␤1 oligonucleotide. Binding activities in cells incubated in 0.3 mM K ϩ for 1 day are expressed relative to the value derived from cells exposed to 5.4 mM K ϩ . Data from five independent experiments done in duplicate were pooled (n ϭ 10). *, p Ͻ 0.05. reporter constructs GC.1 Mut, GC.1,2 Mut, and GC.1,2,3 Mut (Fig. 8A). Wild type and GC box mutant reporter constructs were transiently transfected into neonatal rat cardiac myocytes and the effect of incubation in 0.3 mM K ϩ for 1 day was examined (Fig. 8B). Mutation of the core sequence of the proximal GC box at Ϫ58 to Ϫ56 (GC.1 Mut) decreased the magnitude of the low K ϩ effect by 31%. A 3-base pair mutation of both the proximal GC box and the adjacent putative GC element at Ϫ67 to Ϫ65 (GC.1, 2 Mut) resulted in no further decrease in low K ϩ -mediated ␤1 gene transcription. Interestingly, the stimulatory effect of low K ϩ was abolished with reporter construct GC.1,2,3 Mut in which the core sequence of all three potential GC box motifs was altered. These results indicate that within the native ␤1 proximal promoter two distinct regions are necessary for complete activation of transcription in response to low K ϩ , the proximal GC box at Ϫ59/Ϫ55 and a potential GC element at Ϫ102/Ϫ98. Gel-mobility shift analysis with wild type and mutant oligonucleotides indicated that the Sp family of transcription factors bound the distal GC box (Fig. 9A). By contrast, the putative GC box at Ϫ68/Ϫ64 does not appear to be functional in binding Sp transcription factors. In addition, the functional GC box element at Ϫ59/Ϫ55 appears to demonstrate higher affinity for Sp factors compared with the distal Ϫ102/ Ϫ98 site. Importantly, the distal GC box at Ϫ102/Ϫ98 binds transcription factors Sp1 and Sp3 (Fig. 9B) and binding of these factors is increased in response to low K ϩ (Fig. 9C). DISCUSSION It is well established that extended incubation of mammalian cells in medium containing a low external concentration of K ϩ is associated with an up-regulation of Na,K-ATPase activity, content, and subunit mRNA abundance; however, the underlying mechanism(s) remains to be determined at a molecular level. Previous work in our laboratory (21) demonstrated FIG. 7. Enhanced binding of Sp proteins to the ␤1 gene ؊67 to ؊35-base pair region in response to low K ؉ . Electrophoretic mobility shift assay conditions and probe were the same as described in the legend to Fig. 6. Panel A, the effect of addition of competitor oligonucleotides on upper and lower band formation was evaluated in 1 g of nuclear extracts derived from control cells and cardiac myocytes exposed to 0.3 mM K ϩ for 1 day. Lanes 1 and 2 depict binding in the absence of competitor oligonucleotides, lanes 3 and 4 show the effect of preincubation with the unlabeled Ϫ67/Ϫ35 ␤1 oligonucleotide, lanes 5 and 6 represent factor binding in the presence of an USF-1 oligonucleotide, and lanes 7 and 8 show the effect of addition of a Sp1 oligonucleotide containing a consensus GC box sequence. In these experiments a 100-fold excess of competitor oligonucleotides was used. Panel B, electrophoretic mobility shift assay was performed with 1 g of nuclear extracts (lanes 1 and 2), or with extracts preincubated with 1 g of anti-Sp1 (lanes 3 and 4), anti-Sp3 (lanes 5 and 6), anti-USF-1 (lanes 7 and 8), or rabbit IgG (lanes 9 and 10). Panel C, quantitation of the effect of inclusion of antibodies on upper band formation. Panel D, binding of 1 g of nuclear extracts derived from control cells and cardiac myocytes exposed for 1 day to 0.3 mM K ϩ to either the wild type Ϫ67/Ϫ35 ␤1 oligonucleotide (lanes 1 and 2) or a GC box mutant Ϫ67/Ϫ35 oligonucleotide (lanes 3 and 4). A minimum of three independent electrophoretic mobility shift assays were performed. that stimulation of Na,K-ATPase ␤1 subunit mRNA content in primary cultures of neonatal rat cardiac myocytes exposed to low K ϩ was mediated by the Ϫ102 to ϩ151-base pair region of the ␤1 promoter. In the present study, we first determined that nucleotides between Ϫ102 and Ϫ42 in the ␤1 promoter were sufficient to confer low K ϩ responsiveness to the luciferase reporter gene. The location of the low K ϩ response region within the non-transcribed, 5Ј-flanking region of the Na,K-ATPase ␤1 gene indicates an augmentation of the rate of ␤1 gene transcription is responsible for low K ϩ -mediated in-creased abundance of ␤1 mRNA. This finding is consistent with previous studies in which increased intracellular Na ϩ (28) and exposure to ouabain (29) stimulated Na,K-ATPase ␣1 and ␤1 gene transcription; thereby supporting the contention that elevated intracellular Na ϩ is a common proximal signal in low K ϩ and ouabain-mediated regulation of Na,K-ATPase subunit gene transcription.
To begin to elucidate the molecular signal transduction pathway that leads to up-regulation of Na,K-ATPase ␤1 subunit gene expression, we determined that extracellular Ca 2ϩ was necessary for low K ϩ -induced increases in ␤1 mRNA content and rate of ␤1 gene transcription. There are conflicting reports concerning the requirement for Ca 2ϩ in regulation of Na,K-ATPase subunit gene expression following persistent inhibition of Na,K-ATPase function. For example, a role of increased intracellular Ca 2ϩ in the ouabain-induced, transient augmentation of Na,K-ATPase ␣1 and ␤1 gene transcription in rat kidney tubule cells has been proposed (29). Similarly, Huang et al. (30) found that Na,K-ATPase ␣3 subunit mRNA abundance was repressed by ouabain treatment of neonatal rat cardiac myocytes in a Ca 2ϩ -dependent manner. By contrast, Yamamoto et al. (12) demonstrated that incubation of primary cultures of neonatal rat cardiac myocytes with 1 mM ouabain yielded a transient, Ca 2ϩ -independent increase in ␣1 mRNA content. The discrepancy in the requirement for extracellular Ca 2ϩ in the ouabain-mediated effects on Na,K-ATPase ␣1 and ␣3 mRNA content may reflect an isoform-specific difference in signal transduction pathways. Alternatively, as discussed previously (30), the apparent Ca 2ϩ independence of the ouabainmediated up-regulation of ␣1 mRNA abundance described by Yamamoto et al. (12) might represent a nonphysiological response to a toxic concentration of ouabain.
Our observation that incubation of neonatal rat cardiac myocytes in the presence of 1 mM or lower concentrations of external K ϩ resulted in a stimulation of Na,K-ATPase ␤1 reporter gene expression is consistent with an earlier study which demonstrated an increase in Na,K-ATPase content in cardiac myocytes exposed to 1 mM K ϩ (10). Moreover, the lack of an effect of 2 mM K ϩ on ␤1 promoter activity conforms to the contention that acute Na,K-ATPase inhibition, rather than low K ϩ by itself, is the proximal event in the signal transduction pathway that leads to regulation of Na,K-ATPase expression. The molecular events underlying low K ϩ control of ␤1 promoter activity are therefore likely to be distinct from the mechanism(s) responsible for tissue-specific alterations in Na,K-ATPase subunit gene expression in response to hypokalemia in vivo that yield serum K ϩ concentrations in the range of 2 to 3 mM (24).
In the current study we examined the possibility that either an increase in the intracellular concentration of Ca 2ϩ or an enhanced rate of myocyte contraction might be responsible for the low K ϩ -mediated augmentation in Na,K-ATPase ␤1 gene expression. Our finding of increased ␤1 reporter gene expression in contractile-arrested cardiac myocytes incubated with 54 mM K ϩ is not consistent with the hypothesis that increased contraction is the stimulus which promotes ␤1 up-regulation in response to low K ϩ . Similarly, we provide evidence that elevated intracellular Ca 2ϩ is not solely responsible for the low K ϩ effect on Na,K-ATPase ␤1 gene expression. Specifically, we show that the low K ϩ response region is present within the Ϫ102 to Ϫ42-base pair region of the ␤1 promoter, whereas nucleotides between Ϫ55 to ϩ151 of the ␤1 gene are necessary for increased ␤1 promoter activity in response to elevated intracellular Ca 2ϩ in myocytes cultured in the presence of 54 mM K ϩ . It is important to note, however, that our findings of increased Na,K-ATPase ␤1 gene expression as a result of elevated intracellular Ca 2ϩ are consistent with a reported Ca 2ϩ -FIG. 8. Effect of GC box mutations on low K ؉ -induced transcription of the ␤1 promoter. Panel A, sequence of the ␤1 gene low K ϩ response region with potential GC box motifs underlined. R refers to the reverse complement. Wild type and GC box mutant promoter constructs are shown. Panel B, neonatal rat cardiac myocytes were incubated and transfected as described in the legend to Fig. 1 with 5 g of luciferase plasmids containing mutated core GC box motifs. Cells were exposed to either 5.4 or 0.3 mM K ϩ for 1 day and luciferase activity was assayed. Data shown was derived from three to five independent experiments (n ϭ 8 -10). *, p Ͻ 0.05. dependent stimulation of ␤1 mRNA content in rat kidney outer medullary tubular suspensions (31).
To further investigate the molecular basis of low K ϩ -induced Na,K-ATPase ␤1 gene transcription, we carried out a detailed transient transfection analysis of the effect of 5Ј-flanking region deletions of the ␤1 promoter on low K ϩ responsiveness. This approach lead to the localization of a low K ϩ response element to the Ϫ62 to Ϫ42-base pair region of the ␤1 promoter. Importantly, ␤1 promoter activation by low K ϩ was lost with a deletion construct that disrupted a GC box motif at Ϫ61/Ϫ52. The role of additional potential GC box elements within the Ϫ102 to Ϫ42-base pair region of the ␤1 gene in the low K ϩ response was evaluated in transient transfection studies with luciferase plasmids containing 3-base pair mutations in core FIG. 9. Binding of Sp1 and Sp3 to the distal GC box in the ␤1 gene promoter. Electrophoretic mobility shift assay conditions were the same as described in the legend to Fig. 6. Panel A, the effect of addition of either a competitor GC box oligonucleotide or a GC box mutation on upper and lower band formation was evaluated in 5 g of nuclear extracts derived from control cardiac myocytes incubated in serum-free medium containing 5.4 mM K ϩ . Lanes 1, 4, and 7 depict binding to oligonucleotides containing the Ϫ114/Ϫ82, Ϫ86/Ϫ62, and Ϫ66/Ϫ42 regions of the ␤1 promoter. The effect of preincubation with a 100-fold excess of an oligonucleotide containing a consensus GC box sequence on binding to the three oligonucleotides noted above is shown in lanes 2, 5, and 8. Lanes 3, 6, and 9 represent binding to Ϫ114/Ϫ82, Ϫ86/Ϫ62 and Ϫ66/Ϫ42 oligonucleotides each containing a 3-base pair mutation of the core GCG motif. Panel B, the effect of preincubation with Sp1 and Sp3 antibodies on binding to the Ϫ114/Ϫ82 oligonucleotide was evaluated in 5 g of nuclear extracts derived from control cardiac myocytes incubated as in Panel A. 5 g of nuclear extracts was either incubated in the absence of antibody (lane 1), or with extracts preincubated with 5 g of anti-Sp1 (lane 2) or anti-Sp3 (lane 3). Panel C, binding of 5 g of nuclear extracts derived from control cells (lane 1) and cardiac myocytes exposed for 1 day to 0.3 mM K ϩ (lane 2) to the Ϫ114/ Ϫ82 ␤1 oligonucleotide. Three independent electrophoretic mobility shift assays were performed. GCG motifs at Ϫ58/Ϫ56, Ϫ67/Ϫ65, and Ϫ101/Ϫ99. A reduction in the degree of low K ϩ -mediated stimulation of ␤1 gene transcription demonstrated with the Ϫ58/Ϫ56 mutation supports the contention that the proximal GC box is necessary for the low K ϩ response. It is important to note, however, that mutation of the distal GC box at Ϫ101/Ϫ99 yielded a significant diminution of the low K ϩ -induced increase in ␤1 gene expression. Taken together, these results suggest that a cooperative interaction between two functional GC box elements is required to elicit maximal stimulation of Na,K-ATPase ␤1 gene transcription in neonatal rat cardiac myocytes incubated in medium containing low external K ϩ .
To determine whether GC box-binding proteins were involved in the stimulation of Na,K-ATPase ␤1 gene transcription by low K ϩ , we examined binding of nuclear factors to oligonucleotides containing both the proximal and distal GC boxes in electrophoretic mobility shift assays. Two protein-DNA complexes were detected and determined to increase in nuclear extracts derived from low K ϩ -treated cardiac myocytes. The observation of a low K ϩ -mediated increase in both complexes suggests a molecular mechanism in which a stimulatory effect of nuclear factor binding to the low K ϩ elements, rather than the release of a repressor, is responsible for an enhanced rate of ␤1 gene transcription. Inclusion of antibodies against transcription factors Sp1 and Sp3 in mobility shift assays revealed that both Sp1 and Sp3 directly bound the ␤1 gene low K ϩ response elements. To our knowledge, these findings represent the first definition of the cis-element and transcription factors that are necessary for low K ϩ -mediated transcriptional regulation of a Na,K-ATPase subunit gene. Sp1 and Sp3 proteins are ubiquitously expressed and both recognize GC box motifs to modulate rates of gene transcription in many mammalian cells (32). Typically, Sp1 plays a stimulatory role in basal transcription while Sp3 may either activate or repress transcription (32). In our study, we provide evidence that binding of Sp1and Sp3 to proximal and distal GC boxes in the Ϫ102 to Ϫ42-base pair region of the Na,K-ATPase ␤1 gene leads to a stimulation of ␤1 gene transcription in response to low K ϩ . Whether enhanced binding of Sp1 and Sp3 transcription factors to ␤1 gene GC box elements is a consequence of either increased abundance of these trans-acting proteins or post-translational modification remains to be elucidated in future studies.
It is noteworthy that the recovery of nuclear protein increased 26% in cardiac myocytes incubated for 1 day in medium containing low K ϩ compared with controls. Moreover, we have found a significant increase in total cellular protein and RNA content in neonatal rat cardiac myocytes cultured for 3 days in the presence of low K ϩ (data not shown). These observations suggest that exposure of cardiac myocytes to low K ϩ results in cell hypertrophy. Indeed, our findings are consistent with the report of Huang et al. (33) that describes hypertrophy of neonatal rat cardiac myocytes in response to incubation with ouabain. Interestingly, GC box elements have been implicated in ␣ 1 -adrenergic regulation of atrial natriuretic factor (34,35) and skeletal ␣-actin (36) gene transcription and associated hypertrophy of neonatal rat cardiac myocytes. In addition, the nuclear content of Sp1 and Sp3 were shown to be increased in a murine model of cardiac hypertrophy (37). Taken together, it is tempting to speculate a general role for Sp family transcription factors in mediating specific gene transcription during the development of cardiac hypertrophy.
We contend that the delineation of the intracellular signal transduction pathway associated with low K ϩ -dependent, upregulation of Na,K-ATPase ␤1 subunit gene expression will provide physiologically relevant information since partial inhibition of Na,K-ATPase function by several distinct biomolecules has been implicated in the physiological and pharmaco-logical control of Na,K-ATPase activity. For example, digitalis derivatives are widely used to reduce Na,K-ATPase function in the pharmacological treatment of congestive heart failure and atrial fibrillation (38). Furthermore, increased content of endogenous digitalis-like factors has been demonstrated in several pathophysiological conditions including essential hypertension (39,40), pregnancy-induced pre-eclampsia (41), and acute myocardial infarction (42). In addition, endothelin and interleukin-1 have both been shown to inhibit Na,K-ATPase activity via augmentation of prostaglandin E 2 production (43,44). Finally, phosphorylation of the ␣ subunit of Na,K-ATPase by protein kinase A (45), protein kinase C (46), or Ca 2ϩ /calmodulin-dependent protein kinase (47) decreases Na,K-ATPase activity. It is our contention that up-regulation of Na,K-ATPase ␤1 gene expression in response to acute Na,K-ATPase inhibition represents a highly conserved adaptive mechanism that is responsible for maintaining ion homeostasis in a wide variety of cell types.