Tissue-Specific Expression of the Human Brain Natriuretic Peptide Gene in Cardiac Myocytes
Abstract Brain natriuretic peptide (BNP) is a cardiac hormone constitutively expressed in the adult heart. To identify the cis-acting elements involved in regulation of the human BNP gene, we subcloned the full-length promoter (−1818 to +100) and deletions thereof upstream from a luciferase reporter gene and transiently transfected them into primary cultures of neonatal rat atrial and ventricular myocytes and myocardial fibroblasts. Luciferase activity of the full-length construct was higher in ventricular (39 064±8488 relative light units, n=11) and atrial (11 225±1907, n=17) myocytes than myocardial fibroblasts (329±113, n=5). Maximal promoter activity in ventricular and atrial myocytes was maintained by sequences positioned between −1818 and −1283 relative to the transcription start site. Deletion to −1175 resulted in a decrease, whereas further deletion to −500 effected an increase in reporter activity in both cell types. In ventricular and atrial myocytes, deletion from −500 to −40 reduced luciferase activity 20-fold and 2-fold, respectively, whereas in myocardial fibroblasts, deletion to −40 upregulated the BNP promoter 2-fold. Of note, deleting 16 bp between −127 and −111 reduced luciferase activity 7-fold and 4-fold in ventricular and atrial myocytes, respectively, but had essentially no effect on luciferase activity in fibroblasts. Placement of sequences lying between −127 and −40 upstream from a heterologous thymidine kinase promoter resulted in reporter expression that was 7.4-fold greater than the vector alone in ventricular myocytes, approximately 2-fold greater in atrial myocytes, and equivalent to the vector alone in fibroblasts. For study of activity of the human BNP promoter in adult myocytes, either 408 or 97 bp of 5′ flanking sequence coupled to the luciferase reporter gene was injected into the apex of adult male Sprague-Dawley rat hearts. After 7 days, luciferase activity in the injected myocardium was 9.8-fold higher for the longer construct (70 683±14 744 versus 7223±3920, n=4, P<.01), consistent with our in vitro data. These data indicate that (1) the full-length human BNP promoter is more active in ventricular versus atrial myocytes and essentially inactive in fibroblasts, (2) the distal BNP promoter contains both positive and negative regulatory elements, (3) a region of the proximal BNP promoter located between −127 and −40 confers tissue specificity, and (4) the BNP promoter is active after injection into the adult rat heart.
Originally isolated from porcine brain, BNP, a member of the natriuretic peptide gene family, has since been shown to be synthesized in the heart of a number of species.1 2 In contrast to ANF, BNP is primarily a ventricular hormone that is synthesized and secreted constitutively in the adult.1 2 3 4 Like ANF, BNP has diuretic, natriuretic, and vasorelaxant properties.1 2 Additionally, it may have antigrowth actions in the vasculature operating either directly or indirectly through stimulation of C-type natriuretic peptide synthesis,5 which suppresses mitogenesis in vascular smooth muscle cells in vitro6 and in vivo.7 Pathophysiological conditions, such as left ventricular hypertrophy, hypertension, heart failure, and myocardial infarction, result in enhanced synthesis and secretion of BNP and ANF; in advanced stages of heart failure, plasma BNP levels can exceed those of ANF, suggesting differences in regulation.1 2 Previous studies have also shown differences in posttranscriptional regulation of BNP and ANF.3 A comparison of regulatory regions in the proximal promoter of the hBNP and human ANF genes indicates only 50% sequence similarity,8 9 although both contain sequences with high homology to known regulatory elements, including 12-O-tetradecanoylphorbol-13-acetate (phorbol ester)–responsive elements (TRE or AP-1 binding site), serum response elements (CArG/SRE), and GATA sites. As the BNP gene is constitutively expressed in the adult left ventricle, whereas ANF gene expression is extinguished early in neonatal life, it is likely that the molecular regulation of these two genes differs. To study transcriptional regulation of the hBNP gene, we obtained sequence for 1831 bp of 5′FS extending upstream from the transcription initiation site. Deletions were generated from −1818 to −40 and then subcloned upstream from a luciferase reporter gene. Chimeric genes were transiently transfected into cultured neonatal atrial and ventricular myocytes, myocardial fibroblasts, and intact ventricular myocardium to localize cis-elements conferring cell type–specific expression. These data suggest the presence of both positive and negative regulatory elements that are selectively recognized in different cell types.
All protocols using live animals were approved by the Henry Ford Hospital Committee for Care and Use of Experimental Animals. Atrial-enriched and ventricular myocyte–enriched cultures were generated from Sprague-Dawley neonatal rat pups (Charles River, Kalamazoo, Mich) as described previously.3 Myocytes were separated from myocardial fibroblasts through a differential preplating step. Atrial and ventricular myocytes were plated at a density of 1×105 cells per centimeter squared (1×106 cells per well in a six-well plate) in DMEM (GIBCO) containing 100 U/mL penicillin, 100 μg/mL streptomycin, 2 mmol/L glutamine, and 10% fetal bovine serum (HyClone). On days 3 and 4 after preparation, medium was changed to DMEM supplemented with 5 mg/L insulin and transferrin and 2.5 mg/L selenium. On day 5, cells were removed from the wells and lysed for assay of luciferase and either β-gal or protein.
Myocardial fibroblasts were generated from the preplating step described above. Cells were maintained in DMEM plus 10% fetal bovine serum. Cells were passaged twice and used for transfection studies when confluent (passage 3).
A 3.2-kb fragment of the hBNP gene containing approximately 1.9 kb of 5′FS was generously provided by Dr Gordon Porter (Scios-Nova, Mountain View, Calif). The sequence from −385 through +1323 was published previously.8 The entire 5′FS (−1 to −1831) was sequenced with a combination of automated (Taq dideoxy chain termination fluorescent cycle sequencing, Applied Biosystems) and manual (fmol DNA sequencing system, Promega) sequencing with nine sense and five antisense strand oligonucleotides. The DNA sequence has been submitted to the GenBank database with accession number U34833.
To generate deletions of the 5′FS, we used either convenient restriction sites (eg, BglII, BstXI, and BsaBI for −1595, −1306, and −1283 hBNP, respectively) or PCR. Oligonucleotides included restriction sites at their 5′ borders to facilitate subcloning (HindIII site on sense primers and BamHI on the antisense primer are not included in the following sequences). Sense strand oligonucleotides included −1818 hBNP (5′-GTAGAAACACCTTGTGATCAC-3′), −1175 hBNP (5′-CCAGGCTGGAGTGCAGTGGCG-3′), −906 hBNP (5′-GTTGGCTTGGTGGGGGAGAGG-3′), −500 hBNP (5′-CAGGCAGGGTGCACAGCG-3′), −449 hBNP (5′-AACTCGCGCGGGGAG-3′), −408 hBNP (5′-CTTGGCCGGGGCTGTTTT-3′), −198 hBNP (5′-GCCGACCCGGCCCATTTC-3′), −127 hBNP (5′-GCTCATTCCCGGGCC-3′), −111 hBNP (5′-TGATCTCAGAGGCCC-3′), −97 hBNP (5′-CGGAATGTGGCTGATAAA-3′), −70 hBNP (5′-AGACCTGCATGGCAG-3′), and −40 hBNP (5′-AGCTCCAGGATAAAA-3′). The antisense oligonucleotide was located between +83 and +100 (5′-GGGACTGCGGAGGCTGCT-3′). PCR was performed with the hBNP gene as a template. Standard reagents were obtained from Perkin-Elmer. PCR products of the expected size were cut with HindIII and BamHI and subcloned into digested ▵luciferase vector.10 For −1595, −1306, and −1283 hBNPLuc, restriction digestion was followed by blunt-end ligation of HindIII linkers, HindIII digestion, and religation of the plasmid.
A BamHI–BglII fragment containing a minimal TK promoter (−49 to + 53) was subcloned into identical sites in ▵luciferase to generate TKLuc. Fragments of the proximal BNP promoter (−127 to −40 and −111 to −40) were generated by PCR with sense primers listed above and the antisense primer 5′-TGAGTGTCGGGCCTG-3′ (−40 to −54). PCR-amplified fragments of the correct size were cloned upstream of TKLuc in the HindIII and BamHI sites, and constructs were confirmed by dideoxy sequencing.
Transfection and Luciferase Assay
Freshly isolated atrial and ventricular myocytes were transiently transfected by electroporation at 280 V and 250 μF with a Bio-Rad Gene Pulser. Myocardial fibroblasts at passage 3 were transfected at 250 V and 250 μF. Either cotransfected Rous Sarcoma virus (RSV)–β-gal or protein (Coomassie protein reagent, Pierce Chemical Co) was used to normalize luciferase activity. For ventricular myocytes, 1.25 μg of each of the hBNPLuc deletions was transfected per 3×106 cells. For atrial myocytes and myocardial fibroblasts, 5 μg of the hBNPLuc deletions was transfected per 2×106 to 3×106 cells. After transfection, cells containing each of the hBNPLuc deletions were plated onto 3 wells of a 6-well plate (for ventricular myocytes) or 3 to 4 wells of a 12-well plate (for atrial myocytes and myocardial fibroblasts). Two days after transfection, the medium was changed to serum-free DMEM; cells were harvested 48 hours later (approximately 90 hours after transfection), lysed, and assayed for luciferase (Luciferase Assay System, Promega) and galactosidase (Galacto-Light chemiluminescent assay, Tropix) with an OptoComp 1 luminometer (MGM Instruments) according to the manufacturers’ protocols. Duplicate aliquots of cell lysates from triplicate or quadruplicate wells were assayed and averaged. Luciferase activity (in RLU) for each hBNPLuc deletion was normalized to either β-gal or protein (RLU/β-gal or RLU/mg protein) and then normalized to −500 hBNPLuc or TKLuc (relative luciferase activity). Relative luciferase activity of −500 hBNPLuc or TKLuc was arbitrarily set to 1. At least two separate preparations of each plasmid were used for each experimental group. Data are expressed as mean±SE. Results were analyzed by one-way ANOVA and multiple pairwise comparisons made by the Student-Newman-Keuls method. A value of P<.05 was considered significant.
In Vivo Injection of DNA Into Rat Hearts
Male Sprague-Dawley rats (275 to 300 g) were anesthetized with pentobarbital (50 mg/kg), intubated, and connected to a rodent ventilator (Harvard model 683). A thoracotomy was performed and the heart exposed. With a 30-g, 0.5-in needle attached to a 1-mL disposable syringe, 95 μg of either −408 or −97 hBNPLuc in 100 μL phosphate-buffered saline was injected into the apex and free wall of the left ventricle (day 1). Four rats were injected with each hBNPLuc deletion. Seven days after injection, rats were killed and the hearts removed. The apex and free wall (but not the septum) were dissected from the left ventricle and homogenized by a polytron in 1.2 mL of 10 mmol/L Tris, pH 7.5. After homogenization, 0.3 mL of 5× reporter lysis buffer (Promega) was added. Cellular debris was removed by centrifugation in a microfuge, and the supernatant was assayed for luciferase activity. Homogenization of the noninjected right ventricle or the apex of the left ventricle injected with ▵luciferase plasmid was used as a negative control (<100 RLU/mg protein). A preliminary experiment was done to determine whether to normalize luciferase activity to coinjected RSVCAT (chloramphenicol acetyltransferase) activity versus total protein. We found that both methods gave similar results, and so luciferase activity was expressed as RLU per milligram protein. Differences between luciferase activity (mean±SE, n=4) in −408 hBNPLuc hearts versus −97 hBNPLuc hearts were analyzed by Student’s t test; a value of P<.05 was considered significant.
Fig 1⇓ shows the sequence of the hBNP 5′FS together with the previously published sequences of the rat11 and mouse12 genes. Moderate homology exists within the group; the hBNP gene is 65% and 77% homologous with the rat (r) and mouse (m) BNP genes, respectively. Although only limited sequence is available for the mBNP 5′FS, its homology with the rat gene is 90%, reflecting their close phylogenetic relationship. A number of potential regulatory motifs can be identified in each gene. Of note, a GATA motif, the consensus sequence of which is A/TGATAA/G, is located at −85 (relative to the start site) in the human gene. Two such motifs (−96 and −85) have been shown to be functionally important for expression of the rBNP gene.11 13
Previous studies have shown that BNP mRNA is constitutively expressed in the adult rat right atrium and left ventricle and in neonatal atrial and ventricular myocytes.3 4 Although ventricular expression is approximately 30% to 50% of that seen in atrial myocytes, BNP is considered primarily a ventricular hormone because of the greater mass of the ventricles versus atria.3 4 We first asked whether differences in the levels of BNP mRNA in atrial and ventricular myocytes were due to cell type–specific differences in transcriptional regulation of the gene. To do this, we transfected −1818 hBNPLuc into atrial and ventricular myocytes and assayed luciferase activity as an indirect measurement of transcriptional activity. Fig 2⇓ shows that −1818 hBNP luciferase activity was significantly higher in ventricles versus atria (39 064±8488 RLU/β-gal [n=11] versus 11 225±1907 [n=17]; P<.01), despite the fact that ventricles were transfected with only 25% of the amount of DNA used for the atrial myocytes. As expected, hBNPLuc activity was minimal in myocardial fibroblasts (329±113 RLU/β-gal, n=5), with less than 1% of the activity found in ventricular myocytes. Low expression in fibroblasts predominantly reflected poor expression of the hBNP promoter, although decreased transfection efficiency also played a role because a positive control, pCMVluciferase, was expressed at only 20% of the ventricular level (unpublished observations, 1995). Thus, it appears that although differential transcriptional regulation may account for differences in BNP expression in myocardial versus nonmyocardial cells, it cannot be readily invoked to account for differences in atrial versus ventricular BNP mRNA levels.
We next attempted to localize specific regions of the BNP gene that might confer cell type–specific expression. Transient transfection of ventricular myocytes with −1818, −1595, −1306, or −1283 hBNPLuc resulted in high levels of luciferase activity (Fig 3A⇓). Deletion of an additional 108-bp region containing CTTT repeats (to −1175 hBNPLuc) resulted in a 10-fold drop in luciferase activity, suggesting the presence of a positive regulatory element or structural motif necessary for optimal activity. Further deletion to −500 effected a 3-fold upregulation of luciferase activity, suggesting the elimination of a negative regulatory element. Progressive deletions of the proximal promoter (−500 to −40) generally resulted in a decline in luciferase activity, probably reflecting the systematic elimination of positive regulatory elements lying in close proximity to the promoter. However, deletion from −198 to −127 resulted in an increase in luciferase activity to a level that was roughly 60% that of −1818 hBNPLuc, whereas elimination of a 16-bp region between −127 and −111, containing a putative M-CAT site (a muscle-specific regulatory element with the consensus sequence CATTCCT) and a GC-rich area, reduced activity 7-fold. Luciferase activity of −40 hBNPLuc, containing only the basal promoter, was near background levels (approximately 1100 versus 850 RLU/β-gal for ▵luciferase).
Overall, transfection of the same constructs into atrial myocytes gave a profile similar to that shown for ventricular myocytes (Fig 3B⇑), but the differences were less marked. For instance, in atrial myocytes, deletion from −1595 to −1175 and −127 to −111 each resulted in only a 4-fold decrease in luciferase activity instead of the 10-fold and 7.4-fold decreases (respectively) seen in ventricular myocytes. Progressive deletion of the proximal promoter from −500 to −40 resulted in a modest 2-fold reduction in activity versus 20-fold in ventricular myocytes.
Expression of the hBNPLuc deletion constructs in myocardial fibroblasts was very low compared with that in myocytes, and the pattern of expression was also different (Fig 3C⇑). The most striking difference involved the proximal promoter, where deletion from −500 to −40 doubled luciferase activity. Activity levels of −198, −127, −111, and −40 hBNPLuc were all higher than levels of −500 hBNPLuc. Myocardial fibroblasts also differed from myocytes in that deletion from −111 to −97 resulted in an 11-fold decrease in luciferase activity, suggesting the presence of a positive regulatory element that is important for these non-BNP–expressing cells.
Because −127 hBNPLuc demonstrated a significant recovery of reporter expression in atrial and ventricular myocytes, we next studied whether this region was functional outside the context of the BNP promoter. To do this, we subcloned the −127 to −40 and the −111 to −40 regions upstream of a minimal TK promoter linked to a luciferase reporter gene (TKLuc) and analyzed them after transient transfection (Fig 4⇓). The −111/−40 hBNPTKLuc construct tended to inhibit the TK promoter in all three cell types, but this was statistically significant only in fibroblasts. However, the −127/−40 hBNPTKLuc construct, which contains two M-CAT sites, one AP-1–like site, and one GATA element, upregulated the TK promoter 7.4±2.1-fold in ventricular myocytes and 2±0.57-fold in atrial cells. This fragment was inactive in myocardial fibroblasts.
Because regulation of hBNP gene expression may differ in neonatal versus adult myocytes, we wished to confirm our in vitro data in an in vivo model by examining expression of hBNPLuc constructs after injection into adult rat myocardium. As shown in Fig 5⇓, −408 hBNPLuc was expressed at a high level in adult ventricular myocytes 7 days after injection, consistent with constitutive expression of BNP in the adult heart. The −97 hBNPLuc construct displayed approximately 10% of the activity of the longer construct. Thus, the chimeric hBNPLuc constructs are active in adult ventricular myocytes. Furthermore, as in the in vitro model, expression appears to depend on regulatory sequences present between −408 and the transcription start site.
Our data as well as data from other researchers13 suggest that the BNP gene is preferentially expressed in cardiac myocytes and essentially not expressed in myocardial fibroblasts and that this differential expression derives, at least in part, from selective utilization of the hBNP promoter in myocytes. Both the hBNP and rBNP genes are constitutively expressed in adult atrial and ventricular myocytes, and higher levels of peptide and mRNA are generally observed in the atria.2 3 14 15 Moreover, introduction of chimeric rBNP luciferase reporter gene constructs into neonatal atrial and ventricular myocytes results in similar levels of expression (when normalized to pRSVLuc).13 In contrast, our own data suggest that the hBNP promoter is more active in ventricular versus atrial myocytes. This discrepancy could reflect differences in experimental protocol, intrinsic differences between the rat and human genes, or greater fibroblast contamination of atrial versus ventricular myocytes. Alternatively, it could imply that differences in the steady-state levels of BNP mRNA in atrial and ventricular myocytes arise from posttranscriptional regulatory events. The short half-life of BNP mRNA,3 16 perhaps related to instability motifs present in the 3′ untranslated region,8 makes stabilization of the BNP transcript a potentially important locus for regulatory control.
Analysis of the distal hBNP 5′FS indicates that a number of potential cis-elements with high homology to known regulatory sequences are present in other cardiac-specific genes, including TRE (AP-1), AP-2, myocyte-specific enhancer factor-2 (MEF-2), GATA, M-CAT, and SRE11 13 17 18 19 20 21 22 sites. In the distal 5′FS between −1818 and −1283, there are three GATA elements and an AP-1 site. Our deletion analysis suggests that these sites are not critical for expression of the hBNP gene in ventricular myocytes, as all four deletions within this region gave high levels of expression. The region from −1268 to −1178 is a CT-rich region with 14 CTTT repeats; deletion of this region caused a 10-fold decrease in expression (versus −1595 hBNPLuc). Thus, this region contains a positive regulatory element or elements that appear to be necessary for BNP gene expression. Further deletion of the hBNP gene from −904 to −500 resulted in 4.5-fold upregulation, suggesting the presence of a negative regulatory element or elements in this region.
These same hBNPLuc deletions were also analyzed in atrial myocytes and myocardial fibroblasts. In atrial myocytes, the expression pattern was similar to that in ventricular cells except that the effect of deleting the positive (−1283 to −1175) and negative (−906 to −500) elements was abrogated (3-fold downregulation and 1.4-fold upregulation, respectively). Fibroblasts, despite the low level of expression of all distal deletion constructs, showed evidence for a positive regulatory element between −1818 and −1185 but not the negative regulatory element (−906 to −500) known to be functional in myocytes.
Analysis of deletions of the proximal hBNP promoter (−500 to −40) in the three cell types provided further evidence of ventricular-specific expression of the gene. In ventricular myocytes, deletion from −500 to −198 eliminated putative AP-1 and SRE sites and resulted in a 2.5-fold decrease in hBNPLuc activity. Further deletion to −127 resulted in slight upregulation of activity; however, when an additional 16 bp (to −111) containing an M-CAT–like element and a GC-rich element were eliminated, hBNPLuc activity dropped 7-fold relative to −127 hBNPLuc. Elimination of an AP-1 site (by deleting to −97) and a GATA element (by deleting to −70) resulted in 9- and 15-fold decreases in luciferase activity, respectively (relative to −127 hBNPLuc). Deletion to −40, a region containing only the TATA box-equivalent (GATA at −32), reduced activity 20-fold below −127 hBNPLuc. Thus, it appears that tandemly arranged cis-elements, rather than a single dominant element, are responsible for conferring myocyte-specific expression upon the BNP promoter.
Analysis of the proximal deletion constructs in atrial myocytes gave results similar to those in ventricular cells, except that the decreases in promoter activity after deletion downstream from −127 were not as great (2-fold decrease in activity with deletion from −127 to −40 versus 20-fold in ventricular myocytes). In contrast to myocytes, hBNPLuc expression in fibroblasts showed a very different pattern, with increasing luciferase activity after deletion from −500 to −40. We have previously identified a similar phenomenon in the human ANF gene.9 The −198, −111, and −40 hBNPLuc constructs all gave similar levels of expression. These data suggest that upstream regulatory elements may negatively regulate the proximal promoter. Elimination of these elements uncovers basal promoter activity, which, in fibroblasts, fails to be strongly activated because of the absence of myocyte-specific regulatory factors.
The M-CAT–like element positioned at −124 alluded to above is of particular interest. As shown in Fig 4⇑, when the −127/–40 region of the proximal hBNP promoter was coupled to the TK promoter, it conferred cardiac myocyte–specific expression on the TK promoter. This construct was essentially inactive in fibroblasts. Deletion of a 16-bp region containing the M-CAT–like site abrogated this stimulatory effect in both ventricular and atrial myocytes. Other studies have shown that the M-CAT element, which binds transcriptional enhancer factor-1 (TEF-1), is important in the regulation of the β-myosin heavy chain19 20 and cardiac troponin T21 genes. The −111/−40 region of the hBNP gene, containing one AP-1 and a single GATA element, is inactive in myocytes when coupled to TKLuc and has low activity when coupled to its own promoter. Thus, regulation of the hBNP gene is potentially different from that of the rBNP gene, in which two adjacent GATA elements positioned between −116 and −80 in the proximal promoter have been shown to be critical determinants of activity.11 13
Because BNP gene regulation may differ in neonatal versus adult myocytes, we injected two BNPLuc constructs, which had different activity in neonatal ventricular myocytes, into the left ventricle of adult rats. These studies showed that the hBNP promoter was able to direct cardiac-specific expression in this model, as has been shown previously for regulatory sequences from the α-myosin heavy chain,23 24 creatine kinase M,25 and cardiac troponin C26 genes. In our studies, the difference between −408 hBNPLuc and −97 hBNPLuc activity in adult hearts was twice that of neonatal cells. This may reflect a change in the regulation of the BNP gene with age (neonate versus adult), differences in the transfection methodology, or intrinsic differences in the biology of cultured cells versus intact myocardium. To differentiate between these possibilities and to explore the developmental regulation of the hBNP gene in greater detail, we are developing lines of transgenic mice that express chimeric hBNPLuc genes. Preliminary studies indicate that BNP gene expression is restricted to the heart and a few extracardiac locations (unpublished observations, 1995).
These data provide a detailed analysis of the positive (−1283 to −1175 and −127 to −111) and negative (−906 to −500 and −198 to −127) regulatory elements that contribute to BNP gene expression in cardiac myocytes. We have also delineated a region in the proximal promoter (−127 to −40) that confers cardiac myocyte–specific expression, particularly ventricular myocyte expression, and strongly activates a heterologous promoter. Our data also suggest that regulation of the hBNP and rBNP genes may be different. Deletion of the M-CAT element in the hBNP gene reduces promoter activity to very low levels despite the presence of a GATA element. In the rBNP gene, however, the GATA elements are critical for expression.
Selected Abbreviations and Acronyms
|5′FS||=||5′ flanking sequence(s)|
|ANF||=||atrial natriuretic factor|
|BNP||=||brain natriuretic peptide|
|DMEM||=||Dulbecco’s modified Eagle’s medium|
|hBNP||=||human brain natriuretic peptide|
|PCR||=||polymerase chain reaction|
|RLU||=||relative light unit|
This work was supported by HL-03188 and HL-28982 (M.C.L.) and HL-35753 (D.G.G.). We thank Dr Gordon Porter of Scios-Nova, Mountain View, Calif, for the hBNP gene and Dr Ding Wang for his help with the PCR sequencing. Kim Saracino and Jodi Sitkins provided excellent technical assistance.
Ogawa Y, Nakao K. Brain natriuretic peptide as a cardiac hormone in cardiovascular disorders. In: Laragh JH, Brenner BM, eds. Hypertension: Pathophysiology, Diagnosis, and Management. 2nd ed. New York, NY: Raven Press Publishers; 1995:833-840.
Nazario B, Hu R-M, Pedram A, Prins B, Levin ER. Atrial and brain natriuretic peptides stimulate the production and secretion of C-type natriuretic peptide from bovine aortic endothelial cells. J Clin Invest. 1995;95:1151-1157.
Porter JG, Catalano R, McEnroe G, Lewicki JA, Protter AA. C-type natriuretic peptide inhibits growth factor-dependent DNA synthesis in smooth muscle cells. Am J Physiol. 1992;263:C1001-C1006.
Thuerauf DJ, Hanford DS, Glembotski C. Regulation of rat brain natriuretic peptide gene. J Biol Chem. 1994;269:17772-17775.
Steinhelper ME. Structure, expression, and genomic mapping of the mouse natriuretic peptide type-B gene. Circ Res. 1993;72:984-992.
Grepin C, Dagnino L, Robitaille, Haberstroh L, Antakly T, Nemer M. A hormone-encoding gene identifies a pathway for cardiac but not skeletal muscle gene transcription. Mol Cell Biol. 1994;14:3115-3129.
Mukoyama M, Nakao K, Hosoda K, Suga S-I, Saito Y, Ogawa Y, Shirakami G, Jougasaki M, Obata K, Yasue H, Kambayashi Y, Inouye K, Imura H. Brain natriuretic peptide as a novel cardiac hormone in humans: evidence for an exquisite dual natriuretic peptide system, atrial natriuretic peptide and brain natriuretic peptide. J Clin Invest. 1991;87:1402-1412.
Hanford DS, Thuerauf DJ, Murray SF, Glembotski CC. Brain natriuretic peptide is induced by α-adrenergic agonists as a primary response gene in cultured rat cardiac myocytes. J Biol Chem. 1994;269:26227-26233.
Olson EN. Regulatory mechanisms for skeletal muscle differentiation and their relevance to gene expression in the heart. Trends Cardiovasc Med. 1992;2:163-170.
Shimizu N, Smith G, Izumo S. Both a ubiquitous factor mTEF-1 and a distinct muscle-specific factor bind to the M-CAT motif of the myosin heavy chain β gene. Nucleic Acids Res. 1993;21:4103-4110.
Kariya K-I, Karns LR, Simpson PC. An enhancer core element mediates stimulation of the rat β-myosin heavy chain promoter by an α1-adrenergic agonist and activated β-protein kinase C in hypertrophy of cardiac myocytes. J Biol Chem. 1994;269:3775-3782.
Farrance IKG, 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.
Molkentin JD, Kalvakolanu DV, Markham BE. Transcription factor GATA-4 regulates cardiac muscle-specific expression of the α-myosin heavy chain gene. Mol Cell Biol. 1994;14:4947-4957.
Kitsis RN, Buttrick PM, McNally EM, Kaplan ML, Leinwand LA. Hormonal modulation of a gene injected into rat heart in vivo. Proc Natl Acad Sci U S A. 1991;88:4138-4142.
Ojamaa K, Petrie JF, Balkman C, Hong C, Klein I. Posttranscriptional modification of myosin heavy-chain gene expression in the hypertrophied rat myocardium. Proc Natl Acad Sci U S A. 1994;91:3468-3472.
Vincent CK, Gualberto A, Patel CV, Walsh K. Different regulatory sequences control creatine kinase-M gene expression in directly injected skeletal and cardiac muscle. Mol Cell Biol. 1993;13:1264-1272.
Parmacek MS, Vora AJ, Shen T, Barr E, Jung F, Leiden J. Identification and characterization of a cardiac-specific transcriptional regulatory element in the slow/cardiac troponin C gene. Mol Cell Biol. 1992;12:1967-1976.