Novel Molecular Mechanism of Increased Myocardial Endothelin-1 Expression in the Failing Heart Involving the Transcriptional Factor Hypoxia-Inducible Factor-1α Induced for Impaired Myocardial Energy Metabolism
Background—Hypoxia-inducible factor (HIF)-1α is an important transcriptional factor that activates the gene expression of glycolytic enzymes, which are activated as compensation for impaired β-oxidation of fatty acid in the failing heart. We reported that cardiac endothelin (ET)-1 expression is markedly increased in heart failure. The mechanism, however, is unknown. Because we found an HIF-1α binding site in the 5′-promoter region of the ET-1 gene, we hypothesized that HIF-1α is involved in this mechanism.
Methods and Results—In rat cardiomyocytes, luciferase assay and electrophoretic mobility shift assay showed that HIF-1α transcriptionally activates ET-1 gene expression by direct interaction with the predicted DNA binding site in the 5′-promoter region. HIF-1α mRNA and ET-1 mRNA in the failing heart increased during the aggravation of heart failure in vivo in animal models, ie, rats with myocardial infarction and hamsters with cardiomyopathy. In cultured cardiomyocytes treated with a mitochondrial inhibitor, HIF-1α mRNA and ET-1 mRNA were markedly increased with activated glycolysis, and antisense oligonucleotide for HIF-1α largely inhibited the increased gene expression of ET-1.
Conclusions—The present study revealed a novel molecular mechanism of upregulation of myocardial ET-1 in heart failure, indicating that induction of HIF-1α to stimulate glycolysis as an adaptation in heart failure against impaired energy metabolism alternatively causes an elevation of cardiac ET-1 gene expression as a maladaptation.
Vascular endothelial cells and cardiomyocytes produce endothelin (ET)-1 with various pathophysiological roles.1 2 3 We previously reported that production of ET-1 in the heart was increased in rats with heart failure4 5 and that an ETA receptor antagonist improved the survival rate and hemodynamic parameters.5 Various ET receptor antagonists were reported to be effective in animals with heart failure.1 5 6 7 The mechanism by which ET-1 gene expression is increased in the failing heart, however, is unknown. The purpose of the present study was to answer this question.
From the point of view of energy metabolism, a switch in a principal ATP source from mitochondrial fatty acid oxidation (β-oxidation system) to a glycolytic system may be involved in the pathophysiology of heart failure, because it has been reported that in heart failure, key enzymes in β-oxidation are downregulated, with impaired incorporation of fatty acid as a substrate for β-oxidation.8 9 In the mechanism of the activation of glycolysis, hypoxia-inducible factor (HIF)-1α, a transcription factor, is known to be involved in the increased expression of glycolytic enzymes.10 11 HIF-1α is involved in the responses of cells subjected to hypoxic stress.10 12 It is also known that HIF-1α transcriptionally regulates erythropoietin13 and vascular endothelial growth factor14 by recognizing a DNA element in regulatory regions and forming heterodimerization with arylhydrocarbon-receptor nuclear translocator (ARNT).11 Therefore, in the progression of heart failure, in which impaired energy metabolism may occur, HIF-1α is likely to be involved in the activation of the glycolytic system. Because we found that a HIF-1α recognition site exists in the 5′-promoter region of the ET-1 gene, we hypothesized that HIF-1α is involved in increased myocardial expression of the ET-1 gene in heart failure.
In the present study, we performed 4 series of experiments. Series 1 was to investigate the level of both HIF-1α mRNA and ET-1 mRNA in the failing heart of 2 animal models with heart failure. Series 2 was performed to examine whether direct mitochondrial inhibition causes increases in HIF-1α mRNA and ET-1 mRNA in cultured cardiomyocytes. Series 3 was performed to investigate whether HIF-1α transcriptionally activates ET-1 mRNA expression through binding to the promoter region of the ET-1 gene. Series 4 was to investigate whether direct inhibition of HIF-1α by antisense oligonucleotide affects ET-1 gene expression to reveal that HIF-1α plays a crucial role for ET-1 gene expression.
Experimental Series 1: In Vivo Models of Heart Failure: Rats With Myocardial Infarction and Hamsters With Cardiomyopathy
2. Cardiomyopathic hamsters (CHF146 hamsters) with heart failure: CHF146 cardiomyopathic hamsters were used in the present study.15 CHF148 hamsters, which do not develop heart failure, were used as controls.
Measurement of the Level of mRNA for ET-1 and HIF-1α in the Failing Heart
1. Analysis of gene expression by RT-PCR: RNA isolation and RT-PCR were performed as in our previous reports.6 16 The upstream and downstream gene-specific primers were as follows: HIF-1α sense: 5′-AGTCAGCAACGTGGAAGG-3′; HIF-1α antisense: 5′-GGGAGAAAAGCAAGTCGTG-3′; ET-1 sense: 5′-TCT- TCTCTCTGCTGTTTGTG-3′; ET-1 antisense: 5′-TTAGTTTTCT- TCCCTCCACC-3′; β-actin sense: 5′-GAAGATCCTGACCGAGC- GTG-3′; and β-actin antisense: 5′-CGTACTCCTGCTTGCTGATCC-3′.
Polymerase chain reaction (PCR) was performed with the annealing temperature and required cycles for each template as follows: 56°C and 28 cycles for HIF-1α, 52°C and 26 cycles for ET-1, and 72°C and 24 cycles for β-actin.
Experimental Series 2: Measurement of mRNA for ET-1 and HIF-1α in Cardiomyocytes Treated With a Mitochondrial Inhibitor
1. Isolation and primary culture of rat cardiomyocytes: As demonstrated in our previous report,16 cardiomyocytes were isolated from the hearts of 2-day-old Sprague-Dawley rat neonates and cultured.
2. In vitro study investigating the effects of cobalt chloride (CoCl2), a mitochondrial inhibitor, on gene expression in cardiomyocytes: In the present study, CoCl2 or rotenone (1 μmol/L), a mitochondrial inhibitor,17 18 was used to impair mitochondrial function. Cardiomyocytes were treated with 100 μmol/L CoCl2 for 6 hours.10 19 Reverse transcription (RT)-PCR was performed for the evaluation of gene expression of atrial natriuretic peptide (ANP), ET-1, and HIF-1α. The upstream and downstream gene-specific primers of ANP were as follows: ANP sense: 5′-ATGGGGCTCCTTCTCCATCACC-3′; and ANP antisense: 5′-TCCGCTCTGGGCTCCAATCCTGT-3′. The annealing temperature and required cycles of PCR for ANP were 65°C and 26 cycles.
3. Measurement of 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT) reduction activity in cultured cardiomyocytes: The mitochondrial function of cardiomyocytes was evaluated by MTT assay as previously described.20
4. Measurement of intracellular ATP in cardiomyocytes: Intracellular ATP contents were evaluated with an ATP assay system kit (a firefly luciferase–based method) (Toyo Beanet).
5. Measurement of glucose consumption of cardiomyocytes: The concentration of glucose in the culture medium was measured with a kit by a method including mutarotase and glucose oxidase (Wako Junyaku Co Ltd).
Experimental Series 3: Analysis of the 5′-Flanking Promoter Region of ET-1 Gene in Cardiomyocytes
1. Plasmid construction for HIF-1α and ARNT expression: A full length of rat HIF-1α cDNA was amplified by use of primers in the mouse HIF-1α DNA sequence.21
2. Analysis of the promoter activity of the ET-1 gene by luciferase assay with the coexpression of HIF-1α and ARNT: Photunis control luciferase vector (Nippon Gene Co Ltd) was used for plasmid construction,16 which contains the 5′-promoter region of the ET-1 gene (a reporter vector). The 5′-flanking promoter region of the rat ET-1 gene, ie, −0.75-kb 5′-ET-1 (−750 to +60), was amplified by PCR from rat genomic DNA by primers as follows: −0.75-kb 5′-ET-1 sense: 5′-TCGGATCCTCCCCTGGT- TTGAATC-3′; and −0.75-kb 5′-ET-1 antisense: 5′-GAGGATCCTGTT-TCTGGAGACTCCT-3′.
As a control, we used control luciferase vector (control vector), which did not possess −0.75-kb 5′-ET-1. We performed transfection experiments 5 times, each experiment in triplicate.
3. Transfection of primary cultured cardiomyocytes with cationic reagents: The primary cultured cardiomyocytes were transfected by a cationic reagent, LipofectAmine (Life Technologies), as shown in our previous study.16 Examples of transfection experiments are (a) control vector: cardiomyocytes transfected by control vector; (b) control vector+ARNT: cardiomyocytes transfected by ARNT and control vector; (c) control vector+HIF-1α: cardiomyocytes transfected by HIF-1α and control vector; (d) control vector+HIF-1α+ARNT: cardiomyocytes transfected by HIF-1α, ARNT, and control vector; (e) reporter vector: cardiomyocytes transfected by the reporter vector; (f) reporter vector+ARNT: cardiomyocytes transfected by the reporter vector and ARNT; (g) reporter vector+HIF-1α: cardiomyocytes transfected by the reporter vector and HIF-1α; and (h) reporter vector+HIF-1α+ARNT: cardiomyocytes transfected by the reporter vector, HIF-1α, and ARNT.
After transfection, luciferase activity was measured by a luminometer according to the manufacturer’s instructions (Nippon Gene Co Ltd).
4. Electrophoretic mobility shift assay (EMSA): Primary cultured cardiomyocytes were transfected by HIF-1α expression vector. HIF-1α–expressing rat cardiomyocytes or nontransfected cardiomyocytes as a negative control were used for EMSA. The nuclear extract samples were incubated22 with an α-32P end-radiolabeled 39-bp probe, 5′-TTACGCGTCCGGCTGCACGTTGCCTGTGGGTACGCGTGG-3′, which included HIF-1α binding site, in the presence of a 100 times greater amount of the nonradiolabeled probe or a point-mutated probe (5′-TTACGCGTCCGGCTGCAAAATGCCTGTGGGTACGCGTGG-3′).
Experimental Series 4: Analysis of Endothelin-1 Gene Expression Treated by Antisense Oligonucleotide for HIF-1α
We prepared antisense oligonucleotide for HIF-1α. As a control, scramble oligonucleotide is also prepared as follows: antisense oligonucleotide: CCTCCATGGCGAATCGGTGC; scramble oligonucleotide: ACTCGTACCGCGGCAGTTCG.
The antisense or scramble oligonucleotide (7.5 μmol/L each) was added to primary cultured cardiomyocytes by use of Lipofectin (Life Technologies) and incubated for 6 hours, followed by 48 hours of incubation with ordinary culture medium. Six hours before RNA isolation (ie, 42 to 48 hours of incubation), antisense oligonucleotide, scramble oligonucleotide, or non–Lipofectin-treated cardiomyocytes were treated with 100 μmol/L CoCl2. Total RNA was isolated for RT-PCR.
All data are presented as the mean±SEM. Differences were analyzed by ANOVA with post hoc analysis and the unpaired Student’s t test. The results were considered statistically significant at a value of P<0.05.
Experimental Series 1: HIF-1α Gene Expression in the Heart Is Progressively Increased in Parallel With an Increase in ET-1 Gene Expression in Heart Failure In Vivo
As shown in Table 1⇓, hemodynamic examination revealed that the rats developed heart failure. Figure 1A⇓ shows that HIF-1α mRNA expression increased with increasing ET-1 mRNA in the post–myocardial infarction failing hearts at 3 weeks (P<0.05 versus control, n=6). At 12 weeks, HIF-1α mRNA was further increased (P<0.05 versus control, n=6), with increased ET-1 mRNA in comparison with a control.
As shown in Table 2⇓, at 30 weeks of age, CHF146 hamsters developed heart failure. In this setting, HIF-1α mRNA expression was increased (P<0.05 versus control, n=8) with increased ET-1 mRNA expression (P<0.05 versus control, n=8) (Figure 1B⇑). As shown in Table 2⇓, at 62 weeks of age, CHF146 hamsters showed a marked increase in left ventricular end-diastolic pressure and central venous pressure. In the 62-week-old CHF146 hamsters, the expression of HIF-1α mRNA in the heart progressively increased with increased ET-1 mRNA expression compared with each expression level in 30-week-old CHF146 hamsters (P<0.05 versus 30 weeks of age and control, n=8) (P value calculated by ANOVA with post hoc analysis) (Figure 1B⇑).
Experimental Series 2: Mitochondrial Dysfunction Induced by Chemical Reagents Causes Increases in the Gene Expression of HIF-1α and ET-1 in Cardiomyocytes In Vitro
To further investigate the mechanism by which HIF-1α mRNA and ET-1 mRNA are increased in parallel fashion, we performed an in vitro study with cultured cardiomyocytes with CoCl2, which causes mitochondrial dysfunction.10 17 18 In Figure 2A⇓, the MTT reduction activity level was decreased by 24 hours of CoCl2 (100 μmol/L) treatment to 28% of the control level (n=5). Under this condition, glucose consumption was reciprocally increased to 321% of that in nontreated cardiomyocytes (n=4) (Figure 2B⇓). Figure 2C⇓ shows that the ATP level was slightly decreased by 24 hours of CoCl2 treatment (n=3).
In CoCl2-treated cardiomyocytes, HIF-1α mRNA expression was increased (n=5) (Figure 3A⇓), accompanied by an increase in glucose consumption (Figure 2B⇑). Under this condition, ET-1 mRNA was also markedly increased (n=5) (Figure 3B⇓). The parallel increases in HIF-1α mRNA and ET-1 mRNA expression were associated with an increase in ANP mRNA expression (n=5) (Figure 3C⇓). Furthermore, it was observed that 1 μmol/L rotenone, a mitochondrial complex I inhibitor, also increases HIF-1α mRNA and ET-1 mRNA expression as well as glucose consumption (data not shown).
Experimental Series 3: Promoter Activity of 5′-ET-1 Is Increased by HIF-1α in Primary Cultured Cardiomyocytes
To investigate whether HIF-1α transcriptionally regulates ET-1 gene expression, luciferase analysis using 5′-ET-1 was performed, because 5′-ET-1 possesses the HIF-1α binding site [5′-(A/G)CGTG-3′] (Figure 4A⇓). Figure 4B⇓ showed that the activity increased markedly, by 14.8-fold, with addition of HIF-1α expression vector. Cotransfection with both HIF-1α expression vector and ARNT expression vector increased the promoter activity tremendously, to 30.6-fold, compared with that of the reporter vector alone. This marked enhancement of promoter activity by the cotransfection suggests that HIF-1α recognizes a specific element in the promoter region of the ET-1 gene and transcriptionally activates ET-1 gene expression.
HIF-1α Binds to the Regulatory Region of the ET-1 Gene
To further clarify whether HIF-1α interacts with the HIF-1α binding site in the promoter region of the ET-1 gene, EMSA was performed with 39-bp radiolabeled probe possessing the predicted HIF-1α binding site (n=3) (arrow in Figure 4A⇑). In Figure 5⇓, the nuclear extract from the non–HIF-1α–transfected cardiomyocytes (nontransfectant) does not show a complex with the radiolabeled probe (lane 1). Nuclear extract from HIF-1α transfectant showed the complex, however (lane 2). This complex was not altered by the point-mutated competitor (lane 3). The competitor with HIF-1α binding site, however, inhibited the complex formation (lane 4). This result suggests that HIF-1α binds to the DNA element in 5′-ET-1 and that HIF-1α is involved in the gene expression of ET-1 as a transcriptional factor.
Experimental Series 4: Antisense Oligonucleotide for HIF-1α Largely Prevents a CoCl2-Induced Increase in ET-1 Gene Expression in Cardiomyocytes
As shown in Figure 6⇓, in cardiomyocytes treated for 6 hours with CoCl2, the antisense oligonucleotide for HIF-1α showed a prevention of an increase in gene expression of ET-1 in comparison with CoCl2 alone or scramble oligonucleotide–treated cardiomyocytes. These data provide conclusive evidence that HIF-1α plays an obligatory role for ET-1 induction.
In the present study, we focused on the impaired energy metabolism of the failing heart, with the hypothesis that HIF-1α is responsible for the upregulation of ET-1 gene expression, because it has been reported that β-oxidation is impaired in heart failure with downregulation of the key enzyme genes of β-oxidation.8 It is also reported that glycolysis may alternatively be increased in heart failure in the impairment of β-oxidation.9 Furthermore, it has been reported that HIF-1α is an important transcriptional factor for the induction of glycolytic enzyme genes, because the key enzymes in glycolysis have been shown to possess the HIF-1α binding site in the promoter region.10 11 The present study revealed for the first time that (1) an HIF-1α binding site exists in the 5′-promoter region of the ET-1 gene; (2) HIF-1α directly binds to this promoter region, as demonstrated by EMSA; (3) HIF-1α transcriptionally activates ET-1 gene expression, as demonstrated by luciferase assay; (4) gene expression of HIF-1α and ET-1 in the failing heart progressively increases during the aggravation of heart failure in vivo in 2 models; and (5) antisense oligonucleotide of HIF-1α largely prevents an increase in cardiac ET-1 gene expression. Consequently, the present study suggests a novel molecular mechanism of upregulation of ET-1 expression in heart failure, ie, impairment of cardiac energy metabolism leading to activation of glycolysis is involved in a marked increase in ET-1 gene expression through the transcriptional factor HIF-1α. Therefore, the present study may present a relevant and novel insight to understanding the mechanism of upregulation of cardiac ET-1 gene expression.
These findings were supported by the results of in vitro study, ie, cardiomyocytes treated with a mitochondrial inhibitor causing a switch in energy metabolism from β-oxidation to glycolysis. In the present study, 6 hours of treatment with CoCl2 was performed, because we confirmed that cultured cardiomyocytes were viable under this condition. We also observed that 72 hours of treatment with CoCl2 finally causes apoptosis.19 Therefore, in the present conditions, cardiomyocytes were in a preapoptotic phase. The point we addressed is that cardiomyocytes increase gene expression of not only HIF-1α but also ET-1 in the preapoptotic phase. Furthermore, a demonstration that antisense oligonucleotide for HIF-1α prevents increased ET-1 gene expression provides conclusive evidence that HIF-1α plays an obligatory role in ET-1 gene expression in cardiomyocytes.
It has been reported that some factors, ie, growth factors, vasoactive substances, and mechanical stress, are involved in the increase in ET-1 gene expression in the failing heart.1 In addition to these factors, the present study revealed that enhanced glycolysis regulated by HIF-1α plays a crucial role for induction of ET-1. Because ET-1 is known to increase cardiac muscle contractility,1 ET-1 possesses some adaptive aspect of supporting contractility of the failing heart.1 ET-1 also exerts unfavorable effects, however, such as myocardial hypertrophy.1 3 Therefore, persistent increases in cardiac ET-1 expression in the failing heart have a pathophysiologically maladaptive aspect. Indeed, it has also been reported that long-term treatment with an endothelin receptor antagonist ameliorates the survival and/or hemodynamics of patients and animal models with heart failure.5 6 7 23 24 Therefore, it is suggested that the present study presents a novel concept that this maladaptive response of cardiac ET-1 expression is induced by the adaptive response of HIF-1α for the activation of glycolysis in heart failure.
This study has the following limitations. The in vitro study showed that antisense oligonucleotide for HIF-1α prevented a CoCl2-induced increase in ET-1 gene expression, suggesting that HIF-1α plays an important role in this increase in ET-1. It is possible, however, that another factor is also involved in an increase in ET-1 mRNA in the cultured cardiomyocytes under the present conditions. Furthermore, although gene expression of HIF-1α and ET-1 in the failing heart increased during the aggravation of heart failure in vivo in 2 models, the in vivo study experiments provide only correlative data with respect to HIF-1α and ET-1. In the present in vitro experiments, we directly revealed, by antisense oligonucleotide experiments, the crucial role of HIF-1α in the upregulation of the ET-1 gene. In the in vivo model, however, unlike the in vitro model, it is difficult to perform an experiment that discloses directly whether HIF-1α regulates the transcription of ET-1 gene expression, and therefore, we set this in vivo situation as a study limitation.
In conclusion, the present study revealed a novel molecular mechanism of upregulation of cardiac ET-1 in heart failure: that HIF-1α is profoundly involved in induction of the glycolytic system for the compensation of downregulated β-oxidation of fatty acid; alternatively, as a maladaptive aspect, such an induction of HIF-1α causes elevation of cardiac ET-1 expression, which leads to aggravation of heart failure.
This work was supported by grants-in-aid for scientific research from the Ministry of Education, Science, Sports and Culture of Japan (0006887, 10670629, 11357019, 11557047, 11770346, and 12470147) and by a grant from the Miyauchi project of Tsukuba Advanced Research Alliance (TARA) at the University of Tsukuba.
- Received October 17, 2000.
- Revision received December 31, 2000.
- Accepted January 9, 2001.
- Copyright © 2001 by American Heart Association
Miyauchi T, Yorikane R, Sakai S, et al. Contribution of endogenous endothelin-1 to the progression of cardiopulmonary alterations in rats with monocrotaline-induced pulmonary hypertension. Circ Res. 1993;73:887–897.
Sakai S, Miyauchi T, Sakurai T, et al. Endogenous endothelin-1 participates in the maintenance of cardiac function in rats with congestive heart failure: marked increase in endothelin-1 production in the failing heart. Circulation. 1996;93:1214–1222.
Yamauchi-Kohno R, Miyauchi T, Hoshino T, et al. Role of endothelin in deterioration of heart failure due to cardiomyopathy in hamsters: increase in endothelin-1 production in the heart and beneficial effect of endothelin A antagonist on survival and cardiac function. Circulation. 1999;99:2171–2176.
Spinale FG, Walker JD, Mukherjee R, et al. Concomitant endothelin receptor subtype-A blockade during the progression of pacing-induced congestive heart failure in rabbits: beneficial effects on left ventricular and myocyte function. Circulation. 1997;95:1918–1929.
Sack MN, Rader TA, Park S, et al. Fatty acid oxidization enzyme gene expression is downregulated in the failing heart. Circulation. 1996;94:2837–2842.
Recchia FA, McConnell PI, Bernstein RD, et al. Reduced nitric oxide production and altered myocardial metabolism during the decompensation of pacing-induced heart failure in the conscious dog. Circ Res. 1998;83:969–979.
Semenza GL, Roth PH, Fang HM, et al. Transcriptional regulation of genes encoding glycolytic enzymes by hypoxia-inducible factor 1. J Biol Chem. 1994;269:23757–23763.
Semenza GL. Transcriptional regulation by hypoxia-inducible factor 1. Trends Cardiovasc Med. 1996;6:151–157.
Wang GL, Semenza GL. General involvement of hypoxia-inducible factor 1 in transcriptional response to hypoxia. Proc Natl Acad Sci U S A. 1993;90:4304–4308.
Semenza GL. Regulation of erythropoietin production: new insights into molecular mechanisms of oxygen homeostasis. Hematol Oncol Clin N Am. 1994;8:P863–P884.
Forsythe JA, Jiang BH, Iyer NV, et al. Activation of vascular endothelial growth factor gene transcription by hypoxia-inducible factor 1. Mol Cell Biol. 1996;16:4604–4613.
Haleen SJ, Weishaar RE, Overhiser RW, et al. Effects of quinapril, a new angiotensin converting enzyme inhibitor, on left ventricular failure and survival in the cardiomyopathic hamster. Circ Res. 1991;68:1302–1312.
Weiberg GS. The effect of cobalt ions on energy metabolism in the rat. Can J Biochem Physiol.. 1968;46:549–554.
Weiqiong LW, Markovich R, Chen JW, et al. Regulation of cardiomyocyte apoptotic signaling by insulin-like growth factor I. Circ Res. 1998;83:516–522.
Schreiber E, Matthias P, Muller MM, et al. Rapid detection of octamer binding proteins with “mini-extracts,” prepared from a small number of cells. Nucleic Acids Res. 1989;17:6419.
Packer M, Caspi A, Charlon V, et al. Multicenter, double blind, placebo-controlled study on long term endothelin blockade with bosentan in chronic heart failure: results of the REACH-1 trial. Circulation. 1998;98(suppl I):I-3. Abstract.
Sutsch S, Kiowski W, Yan XW, et al. Short-term oral endothelin-receptor antagonist therapy in conventionally treated patients with symptomatic severe chronic heart failure. Circulation. 1998;98:2262–2268.