Atrial natriuretic peptide (ANP) is a potent diuretic, natriuretic, and vasorelaxant hormone that is expressed early in ventricular hypertrophy. Expression of human ANP is controlled by a series of regulatory elements located in the 5′ flanking sequence of its gene. We generated a series of 5′ deletion mutations extending from −2600 to −1150 relative to the transcription start site and linked them to a chloramphenicol acetyltransferase reporter gene. Using transient transfection analysis, we have identified a negative regulatory element between −1206 and −1152 relative to the start site. Each of a series of 5′ deletion mutants, when introduced into fibroblast cultures, expressed the reporter function at a level that was significantly less (<20%) than that seen with the −1152 reporter construct, whereas comparably transfected atrial cardiocytes demonstrated no change in reporter activity, implying that the repressor function is specific to cell type. The critical region (from −1206 to −1152) associates with a soluble protein present in cardiac fibroblast extracts in a sequence-specific fashion. Deoxyribonuclease I footprint analysis demonstrated the presence of several protected regions, including one that overlies an E-box motif (CAACTG), an element that in other systems has been implicated in promoting differentiation in the myocyte lineage. Site-directed mutagenesis of the E-box motif suppressed both the protein-binding and inhibitory activities of the 54-bp fragment. In summary, we have found a region in the 5′ flanking sequence of the human ANP gene that represses transcriptional activity in nonmyocardial cells. This element may play an important role in the restriction of ANP gene expression to cardiac myocytes.
The definition of factors governing tissue- or cell-specific expression of individual gene products remains one of the elusive problems in eukaryotic cell biology. The current knowledge suggests that such specificity is conferred through the interaction of soluble cellular (or more specifically, nuclear) proteins with one or more key DNA elements in the regulatory regions (usually, but not always, in the 5′ flanking sequence of the target gene). The collective activity of these positive (enhancer) elements and negative (silencer) elements establishes the level of transcriptional activity within a given cell type.
A number of positive regulatory elements have been described in the 5′ flanking sequence of genes expressed in the cardiac myocyte. These include the transcription enhancer factor-1 binding site (M-CAT),1 serum response element/CC(A+T-rich) GGmotif (SRE/CArG),2 myocyte enhancer factor (MEF),3 homeobox,4 and E-box5 6 motifs. The latter (ie, E-box) is the regulatory element originally described by Weintraub and colleagues7 that associates with myoD and the other HLH proteins and is thought to be involved in establishing the muscle phenotype in skeletal myocytes. Although myoD and its close relatives are absent in cardiac myocytes, a number of other basic HLH proteins have been identified in these cells.5 6 7 Some of these appear to have regulatory activity, but none possess the capability of myoD to independently promote the acquisition of a myocyte phenotype.8 In fact, recent cell hybrid studies suggest that there is no dominant factor in cardiac myocytes capable of establishing the cardiac phenotype. Fusion of fibroblasts harboring a cardiocyte-specific transgene with cardiac myocytes failed to activate the transgene in fibroblast-derived nuclei of the heterokaryons.9 In addition to excluding the presence of a myoD equivalent in cardiac myocytes, these latter studies raise the interesting possibility of selective negative regulation of cardiac-specific genes in nonmyocardial cells.
The ANP gene is expressed predominantly although not exclusively in the myocardium. Expression is constitutively high in the atria of rodents and higher mammals, whereas in the cardiac ventricles, expression is more heavily regulated by developmental, neurohormonal, and hemodynamic signals.10 Expression of the endogenous gene is absent in nonmyocardial cells (predominantly fibroblasts) of the rat heart. We have shown previously that this reflects the presence of repressor activity in the distal (−2593 to −1152 relative to the transcription start site) and proximal (−222 to the transcription start site) hANP promoter.11 In the present study, we have dissected this upstream repressor in greater detail and have found that it depends heavily on the presence of an E-box motif in this portion of the gene. In addition, we have identified a soluble fibroblast nuclear protein that associates with this element in a highly sequence-specific fashion.
Restriction and DNA modification enzymes were purchased from Boehringer Mannheim. [γ-32P]ATP (6000 Ci/mmol) and [3H]acetyl coenzyme A (200 mCi/mmol) were purchased from DuPont-NEN Research Products. [α-35S]dATP (1000 Ci/mmol) and Sculptor in vitro mutagenesis kits were purchased from Amersham. Deoxyribonuclease (DNAse) I was obtained from Worthington Biochemicals. Antisera against myoD and myogenin were obtained from C. Ordahl, and those against Pan-1 and Pan-2 from M. German. The antisera against the truncated HLH protein Id was purchased from Santa Cruz Biotechnology, Inc. Other reagents were purchased from standard commercial suppliers and represent the best grades commercially available.
For the sake of convenience, each of the 5′ truncated mutant constructs was defined by the length of the 5′ flanking sequence from the hANP gene linked to the bacterial CAT coding sequence. The −1383, −1349, −1281, −1226, and −1206 hANP CAT 5′ flanking sequence deletion mutant reporter plasmids were generated from −2593 hANP CAT11 by polymerase chain reaction with upstream sense oligonucleotides (for −1383 CAT, 5′-CCAGATCTAGACATGCCACCACACCCAGCTA-3′; for −1349 CAT, 5′-CCAGATCTAGAAGTAGAGACAGAGTTTTGCCA-3′; for −1281 CAT, 5′-CCAGATCTAGACCTTAGCCTCCCAAAGTGCTGGG-3′; for −1226 CAT, 5′-CCTCTAGAAACATTCTTTCTCATTTTACAC-3′; and for −1206 CAT, 5′-CCAGATCTAGAACACAAGGAAAGTAAGGCTGC-3′), each of which incorporated an Xba I site at its 5′ terminus and a downstream antisense oligonucleotide (5′-TATCAACGGTGGTATATCC-3′) derived from CAT coding sequence. The polymerase chain reaction products of appropriate size were restricted with Xba I and HindIII and recloned into compatible sites of pSVoL CAT.11 The −1488 hANP CAT, −1152 hANP CAT, and lamin CAT (promoterless control) constructs are described elsewhere.11 12 The −1226 hANP CAT plasmid was mutated by the Olsen-Eckstein method of site-directed mutagenesis13 with an Amersham in vitro mutagenesis kit. Sequence of the oligonucleotides (read 5′ to 3′) used for the mutagenesis were as follows: M1, TCATTTTACACAaccttccctTAAGGCTGCGAA; M2, GGAAAGTAAGGCgtatcctctGTATGCAACTGG; M3, TGCGAAGAGGTAgtaccagttTTTGTTTTGGGC; and M4, AACTGGTTTGTTggtttagcaAGTACTGGTCTC. The wild-type sequence is indicated by conventional nucleotide nomenclature in uppercase letters and the mutated sequence in lowercase letters. Sequences of all deletion and site-specific mutants were confirmed by dideoxy DNA sequence analysis.
Cell Preparation and Transfection
All studies using animals were in accordance with institutional guidelines at the University of California at San Francisco. Primary cultures of cardiac atrial myocytes were isolated from 1-day-old Sprague-Dawley rat hearts by alternate cycles of trypsin digestion and mechanical disruption as reported earlier.11 Nonmyocytes (primarily fibroblasts) were collected at a differential plating step before removal of myocytes for culture. Cardiac fibroblasts were expanded in culture over a 2-week period before use to reduce the potential for contamination with myocyte elements. Transient transfection was carried out by electroporation at 280 V and 250 μF for the myocytes and 280 V and 960 μF for the fibroblast cells as described previously.11 Variability in transfection efficiency, assessed independently with a Rous sarcoma virus-β galactosidase reporter, was less than 15% within a given experiment. After transfection, cells were plated in 60-mm dishes at a density of 2×106 cells per dish. Cells were harvested and lysed 72 hours after transfection. Protein concentration of each extract was measured with Coomassie protein reagent (Pierce Biochemicals). Cellular protein (100 μg) was used for measurement of CAT activity for each sample. Assays were carried out as described by Neumann et al.14
Gel Shift Assay
Myocytes and nonmyocardial cell monolayers were harvested after 5 days in culture. Nuclei were isolated and extracted as described by West et al.15 In separate experiments, 54-bp (position −1206 to −1152) and 100-bp (position −1226 to −1126) fragments of the hANP gene were end-labeled with [γ-32P]ATP and used as probes in gel mobility shift assays. Five, 15, and 30 μg of nuclear extract were mixed with 1.5×104 cpm (approximately 3 ng) of probe, 2.5 μL of 10× binding buffer (1×=1.0 mmol/L HEPES [pH 7.4], 0.025 mmol/L EDTA, 0.05 mmol/L MgCl2, 6.0 mmol/L KCl, 0.5% glycerol, 0.01% nonylphenoxy polyethoxy ethanol [NP-40], and 0.01% methyl β-d-thiogalactoside), 2 μg of a deoxyinosine-deoxycytidine heteropolymer [Poly(dI-dC), Pharmacia-LKB Biotechnology Inc], other competitors (unlabeled 54-bp or 100-bp hANP gene fragments [positions same as indicated above] or an unlabeled 30-bp DNA fragment derived from the herpes simplex thymidine kinase gene11 16 [positions −53 to −28]), and water to a total volume of 25 μL. The reaction was incubated at room temperature for 35 minutes and then loaded directly onto a nondenaturing polyacrylamide gel in 1× TEA running buffer (6.7 mmol/L Trizma base [pH 7.5], 1 mmol/L EDTA, 3.3 mmol/L sodium acetate). The electrophoresis was performed at constant voltage (8.5 V/cm) with buffer recirculation at 4°C. Gels were then dried and subjected to autoradiography overnight at −70°C. Those studies using specific polyclonal antisera were carried out as described above except that the individual antisera at the indicated dilutions were included in the binding reaction.
DNase I Footprint Assay
For the DNase I footprint assay, 10 ng of a 129-bp hANP genomic fragment (from −1281 to −1152 relative to the transcription start site) was labeled with [γ-32P]ATP at the upstream 5′ terminus and used as a probe to react for 35 minutes at room temperature with 80 or 160 μg of fibroblast nuclear extract in the presence of 20 μL assay buffer (10 mmol/L Tris-HCl [pH 7.0], 2.5 mmol/L MgCl2, 1.0 mmol/L CaCl2, 0.1 mmol/L EDTA, 200 mmol/L KCl, 100 μg/mL bovine serum albumin, and 2 μg/mL calf thymus DNA), 2 μg Poly(dI-dC) as a nonspecific competitor, and water to a total volume of 200 μL. The reaction mix was digested with the indicated concentration of DNase I (0.05 to 0.1 μg/200 μL) at room temperature for 2 minutes. The digestion was terminated by addition of 700 μL stop solution (645 μL 100% ethanol, 5 μL yeast tRNA [1 mg/mL], 50 μL saturated ammonium acetate), followed by DNA precipitation. DNA was resuspended in 5 μL formamide loading buffer (95% formamide, 20 mmol/L EDTA, 0.05% bromphenol blue, 0.05% xylene cyanol), denatured, and loaded on a 6.4% preelectrophoresed polyacrylamide gel containing 8.3 mol/L urea. The sequence ladder was generated by the chain termination method17 with an oligonucleotide primer (5′-CCTTAGCCTCCCAAAGTGCTGGG-3′) whose 5′ terminus corresponds to position −1281.
Our previous investigation identified a negative regulatory element that was operative in cardiac fibroblasts and cultured GC (pituitary tumor) cells but not in atrial or ventricular myocytes.11 In an effort to localize the negative element in greater detail, we generated a series of 5′ deletions extending between −2593 and −1152, linked them to the CAT reporter, and introduced them into atrial myocytes or cardiac fibroblasts isolated from the same neonatal hearts. Deletion mutants upstream of −1383 (ie, those with 5′ termini at −2593, −2203, −1812, and −1488) behaved identically to −1383 CAT (data not shown). As shown in Fig 1⇓, deletions beginning at −1383 and extending to −1152 had little effect on the level of reporter activity in atrial myocytes; however, in the cardiac fibroblasts, the expression of each was relatively low until the deletion reached the −1152 position. These data suggest that a repressor element is located within the 54 bp of sequence separating positions −1206 and −1152 in the hANP gene.
As discussed above, most regulatory elements (positive or negative) seem to accrue their activity as a result of interactions with nuclear proteins present in the target cell. To explore this possibility, we used a 54-bp fragment of the hANP gene spanning the area of interest (−1206 to −1152) in a gel mobility shift assay to probe for the presence of proteins capable of associating with specific DNA sequences therein. As shown in Fig 2⇓, a protein or proteins present in fibroblast nuclear extract bound to this fragment in both a concentration-dependent and sequence-specific fashion. Competition with unlabeled probe, but not with an unrelated DNA fragment, eliminated the DNA-protein interaction. Subsequent DNAse I footprint analysis of this same interaction (Fig 3⇓) demonstrated several protected regions. Of interest, one of these regions overlies an E-box motif (positions −1169 to −1174) similar to those that have been implicated in regulating the transcriptional activity of skeletal and cardiac muscle genes in other systems.5 6
To explore the potential role of this E-box motif in suppressing hANP gene expression in cardiac fibroblasts, we selectively mutated several regions lying between −1152 and −1206 in the −1206 hANP CAT reporter and examined their activity in transiently transfected cardiac fibroblasts. As shown in Fig 4⇓, the highest level of reporter activity was within the group, M3, that harbored the mutation across the E-box motif. This activity approached, though it did not equal, that seen with −1152 hANP CAT. In addition, M1, which harbors a mutation farther upstream, also effected a partial restoration of reporter activity. Taken together, these findings suggest that the E-box motif contributes substantially to the inhibitory activity of the repressor element lying between −1206 and −1152 and that the full effect likely requires additional sequence present in the same region.
Next, we investigated the activity of these mutated sequences with respect to their ability to associate with the fibroblast nuclear binding protein identified in Fig 2⇑. As shown in Fig 5A⇓, the wild-type sequence and three of the mutants (M1, M2, and M4) associated with the protein in the expected fashion. M3, however, displayed no ability to bind to the protein. In a parallel experiment, shown in Fig 5B⇓, unlabeled wild-type fragment competed effectively for binding to labeled wild-type sequence, whereas neither unlabeled M3 nor an unrelated DNA fragment was an effective competitor. Taken together, these data suggest that the predominant DNA protein interaction that is identified through the gel mobility shift assay operates through the E-box motif. Attempts to perturb this complex with antisera directed against the common E-box binding proteins myoD, Id, Pan-1, and Pan-2 were unsuccessful (Fig 6⇓), suggesting that the protein or proteins identified in the complexes above may include a less frequently encountered member or members of this extended regulatory protein family.
A number of different regulatory elements have been described that appear to be important for the tissue-specific expression of genes in cardiac myocytes, and increasingly, it appears unlikely that a single regulatory element or nuclear protein will function as the solitary determinant of cardiac-specific gene expression. Rather, it seems probable that combinations of positive regulatory elements (eg, GATA, M-CAT, SRE/CArG, MEF, stimulatory protein-1 [SP-1], etc) operate in concert to direct transcriptional activity unique to the cardiac myocyte. This scenario appears to be applicable to the ANP gene, in which a number of regulatory elements have been described but none have been ascribed a dominant role in controlling its expression.11 18 19
The nature of negative transcriptional regulation (eg, that responsible for excluding ANP gene expression from nonmyocytes in the heart) is even more unsettled. A number of general mechanisms for negative regulation20 have been proposed, including primary occupancy of a negative regulatory element by a protein or proteins unique to nonexpressing cells, titration of activator protein activity in solution (ie, squelching), and displacement of or interference with a closely positioned DNA-protein complex that positively regulates transcription. The latter two models, by inference, involve interference with the positive regulatory circuitry.
The involvement of an E-box motif is particularly interesting in that similar motifs are known to be important in the regulation of cardiac-specific genes.5 6 Clearly, the motif alluded to here is not critical for ANP gene expression in cultured atrial myocytes (see Fig 1⇑); however, as shown with other E-box–containing promoters,21 these elements can play important modulatory roles with regard to transcriptional regulation that might not be obvious with the simple deletion approach used here. If this E-box motif plays such a role in hANP gene expression, then occupancy of the site by different proteins intrinsic to cardiac myocytes versus nonmyocytes could well provide the switching mechanism that either promotes or suppresses expression of the ANP gene. Genetta et al22 have described a zinc-finger protein, which they term ZEB, that binds to an E-box motif within the immunoglobulin heavy chain (IgH) enhancer and silences its expression. Activation of the gene upon overexpression of the basic HLH protein E2A (an E-box binding protein) is accompanied by displacement of the ZEB repressor from its binding site. Thus, there is some precedent for involvement of an E-box–dependent switching mechanism in tissue-specific gene expression. In fact, preliminary studies suggest that proteins present in cardiocyte nuclear extracts are capable of binding to this same E-box–containing fragment (data not shown), supporting the differential protein binding model.
The functional analyses suggest a potential contribution of sequence upstream from the E-box motif in promoting the inhibitory activity. The fact that the M1 region was not identified as a definitive target for protein binding (Fig 5⇑) despite evidence for functional activity (Fig 4⇑) suggests that DNA-protein contacts in this region, if they exist, are of insufficient affinity to permit identification in the in vitro assays used here. The existence of such interactions is supported, to some degree, by the protection of sequence overlying the M1 region in the DNAse footprint analysis.
The nature of the fibroblast protein associating with the E-box motif remains unknown. Id, a truncated HLH protein, heterodimerizes with and suppresses the activity of several basic HLH proteins involved in tissue-specific transcription and cellular differentiation.23 24 However, Id typically blocks association of these HLH proteins with DNA, a property that is not compatible with participation in the complex identified here. In fact, a specific antisera directed against Id failed to retard or otherwise perturb the fibroblast protein–E-box interaction. Furthermore, overexpression of Id has not been shown to have a significant effect on the expression of a rat ANP-luciferase reporter8 (pANF3003Luc), and Id, if anything, appears to be activated in myocardial cells by the same stimuli that increase ANP gene expression.8 25 Similar findings were obtained after coincubation of the DNA-protein complexes with antisera directed against myoD, Pan-1, and Pan-2. It is possible that the protein identified here is related to the ZEB protein described above or, alternatively, to another HLH protein not detected with the antisera used. Additional studies will be directed toward the identification of the fibroblast and cardiocyte E-box binding proteins associated with this upstream element and elucidation of the mechanism or mechanisms whereby these interactions result in tissue-specific expression of the ANP gene.
Selected Abbreviations and Acronyms
|ANP||=||atrial natriuretic peptide|
|hANP||=||human atrial natriuretic peptide|
This work was supported by grant HL-35753 from the National Institutes of Health. The authors are grateful to Karl Nakamura for assistance with cell preparation and to Fred Schaufele for helpful discussions regarding the mutagenesis protocols. We also wish to express our gratitude to Charles Ordahl, Brian Williams, and Michael German for providing antisera used in these studies.
- Received September 19, 1995.
- Revision received November 7, 1995.
- Revision received April 4, 1996.
Farrance IK, Mar JH, Ordahl CP. M-CAT binding factor is related to the SV40 enhancer binding factor, TEF-1. J Biol Chem. 1992;267:17234-17240.
Sartorelli V, Webster KA, Kedes L. Muscle-specific expression of the cardiac alpha-actin gene requires MyoD1, CArG-box binding factor, and Sp1. Genes Dev. 1990;4:1811-1822.
Zhu H, Garcia AV, Ross RS, Evans SM, Chien KR. A conserved 28-base-pair element (HF-1) in the rat cardiac myosin light-chain-2 gene confers cardiac-specific and alpha-adrenergic-inducible expression in cultured neonatal rat myocardial cells. Mol Cell Biol. 1991;11:2273-2281.
Komuro I, Izumo S. Csx: a murine homeobox-containing gene specifically expressed in the developing heart. Proc Natl Acad Sci U S A. 1993;90:8145-8149.
Gupta MP, Gupta M, Zak R. An E-box/M-CAT hybrid motif and cognate binding protein(s) regulate the basal muscle-specific and cAMP-inducible expression of the rat cardiac alpha-myosin heavy chain gene. J Biol Chem. 1994;269:29677-29687.
Navankasattusas S, Sawadogo M, van Bilsen M, Dang CV, Chien KR. The basic helix-loop-helix protein upstream stimulating factor regulates the cardiac ventricular myosin light-chain 2 gene via independent cis regulatory elements. Mol Cell Biol. 1994;14:7331-7339.
Weintraub H, Davis R, Tapscott S, Thayer M, Krause M, Benezra R, Blackwell TK, Turner D, Rupp R, Hollenberg S, Zhuang Y, Lassar A. The myoD gene family: nodal point during specification of the muscle cell lineage. Science. 1991;251:761-766.
Evans SM, Walsh BA, Newton CB, Thorburn JS, Gardner PD, van Bilsen M. Potential role of helix-loop-helix proteins in cardiac gene expression. Circ Res. 1993;73:569-578.
Evans SM, Tai LJ, Tan VP, Newton CB, Chien KR. Heterokaryons of cardiac myocytes and fibroblasts reveal the lack of dominance of the cardiac muscle phenotype. Mol Cell Biol. 1994;14:4269-4279.
Wu J, Garami M, Cao L, Li Q, Gardner DG. 1,25(OH)2D3 suppresses expression and secretion of atrial natriuretic peptide from cardiac myocytes. Am J Physiol. 1995;268:E1108-E1113.
Olsen DB, Eckstein F. High-efficiency oligonucleotide-directed plasmid mutagenesis. Proc Natl Acad Sci U S A. 1990;87:1451-1455.
Neumann JR, Morency CA, Russian KO. A novel rapid assay for chloramphenicol acetyltransferase gene expression. Biotechniques. 1987;5:444-447.
West BL, Catanzaro DF, Mellon SH, Cattini PA, Baxter JD, Reudelhuber TL. Interaction of a tissue-specific factor with an essential rat growth hormone gene promoter element. Mol Cell Biol. 1987;7:1193-1197.
McKnight SL, Kingsbury R. Transcriptional control signals of a eukaryotic protein-coding gene. Science. 1982;217:316-324.
Sanger F, Nicklen S, Coulson AR. DNA sequencing with chain-terminating inhibitors. Proc Natl Acad Sci U S A. 1977;74:5463-5467.
McBride K, Robitaille L, Tremblay S, Argentin S, Nemer M. fos/jun repression of cardiac-specific transcription in quiescent and growth-stimulated myocytes is targeted at a tissue-specific cis element. Mol Cell Biol. 1993;13:600-612.
Argentin S, Ardati A, Tremblay S, Lihrmann I, Robitaille L, Drouin J, Nemer M. Developmental stage-specific regulation of atrial natriuretic factor gene transcription in cardiac cells. Mol Cell Biol. 1994;14:777-790.
Biben C, Kirschbaum BJ, Garner I, Buckingham M. Novel muscle-specific enhancer sequences upstream of the cardiac actin gene. Mol Cell Biol. 1994;14:3504-3513.
Genetta T, Ruezinsky D, Kadesch T. Displacement of an E-box-binding repressor by basic helix-loop-helix proteins: implications for B-cell specificity of the immunoglobulin heavy-chain enhancer. Mol Cell Biol. 1994;14:6153-6163.
Kreider BL, Benezra R, Rovera G, Kadesch T. Inhibition of myeloid differentiation by the helix-loop-helix protein Id. Science. 1992;255:1700-1702.
Springhorn JP, Ellingsen O, Berger HJ, Kelly RA, Smith TW. Transcriptional regulation in cardiac muscle: coordinate expression of Id with a neonatal phenotype during development and following a hypertrophic stimulus in adult rat ventricular myocytes in vitro. J Biol Chem. 1992;267:14360-14365.