Author + information
- Received October 24, 2006
- Revision received February 28, 2007
- Accepted March 20, 2007
- Published online July 31, 2007.
- Shungo Hikoso, MD, PhD⁎,1,2,
- Yasuhiro Ikeda, MD, PhD§,2,
- Osamu Yamaguchi, MD, PhD⁎,1,
- Toshihiro Takeda, MD, PhD⁎,
- Yoshiharu Higuchi, MD, PhD⁎,
- Shinichi Hirotani, MD, PhD⁎,
- Kazunori Kashiwase, MD, PhD⁎,
- Michio Yamada, MD§,
- Michio Asahi, MD, PhD†,
- Yasushi Matsumura, MD, PhD‡,
- Kazuhiko Nishida, MD, PhD⁎,
- Masunori Matsuzaki, MD, PhD, FACC§∥,
- Masatsugu Hori, MD, PhD, FACC⁎ and
- Kinya Otsu, MD, PhD⁎,⁎ ()
- ↵⁎Reprint requests and correspondence:
Dr. Kinya Otsu, 2-2 Yamadaoka, Suita, Osaka 565-0871, Japan.
Objectives We examined whether the inhibition of apoptosis signal-regulating kinase 1 (ASK1) would attenuate the progression of heart failure in TO-2 hamsters with hereditary dilated cardiomyopathy.
Background Heart failure remains the leading cause of mortality and requires novel therapies targeting the biologically relevant processes within cardiomyocytes that lead to cell death. Apoptosis signal-regulating kinase 1 is a key signaling molecule for cardiomyocyte death.
Methods We generated recombinant adeno-associated virus (rAAV) expressing an N-terminal truncated form of the dominant-negative mutant of ASK1 (ASKΔN(KR)). TO-2 hamsters were subjected to an in vivo rAAV transcoronary transfer.
Results ASKΔN(KR) retained its dominant-negative activity in vitro. The rAAV expressing ASKΔN(KR) treatment inhibited ASK1 activation in the hamster hearts and suppressed progression of ventricular remodeling such as chamber dilation, impairment of contractile and relaxation functions, and fibrosis. Inhibition of ASK1 reduced the number of apoptotic cells and selectively attenuated c-Jun NH2-terminal kinase activation. Although the deficiency of δ-sarcoglycan, a genetic defect in the hamster, leads to the degradation of dystrophin, the treatment significantly protected hearts from this degradation, probably by inhibiting calpain activation.
Conclusions Apoptosis signal-regulating kinase 1 is involved in the pathogenesis of heart failure progression, mediated through c-Jun NH2-terminal kinase-mediated apoptosis and calpain-dependent dystrophin cleavage, and may be a therapeutic target to treat patients with heart failure.
Left ventricular (LV) remodeling is generally accepted as a determinant of the clinical course of heart failure (1). Neurohumoral factors and inflammatory cytokines, which stimulate membrane-bound receptors and downstream multiple cytoplasmic signal transduction cascades, play an important role in the onset and progression of ventricular remodeling (1,2). Despite advances in pharmacological treatments targeting the membrane-bound receptors, prognosis for heart failure, especially severe heart failure, remains poor. A novel and effective drug targeting the cytoplasmic signal transduction cascades that lead to unfavorable cardiac remodeling is necessary to treat patients with heart failure.
Apoptosis signal-regulating kinase 1 (ASK1) is a mitogen-activated protein (MAP) kinase kinase kinase that activates MAP kinase kinase (MKK) 4/7-c-Jun NH2-terminal kinase (JNK) and MKK3/6-p38 MAP kinase (3,4). We have previously reported that ASK1 is activated in pressure overloaded and postinfarct mouse hearts (5) and that the activation leads to not only apoptosis but also necrosis of cardiomyocytes depending on the stress conditions (5,6). Apoptosis signal-regulating kinase 1 knockout mice showed attenuated cardiac remodeling, suggesting that inhibition of ASK1 is beneficial for preventing heart failure (5). However, it remains unclear whether ASK1 inhibition after onset of the disease is therapeutically efficacious for preventing progression of cardiac remodeling and heart failure.
The TO-2 cardiomyopathic hamster is a widely used experimental model of progressive heart failure. The hamster possesses a mutation in the δ-sarcoglycan gene, a component of dystrophin associated protein complex (7), which also has been found in some human patients with dilated cardiomyopathy (8). Deficiency of δ-sarcoglycan causes a concomitant loss of the other sarcoglycan subunits (α, β, γ) and reduced expression level of the dystrophin-glycoprotein complex, leading to mechanical instability of the sarcolemmal membrane in the heart and skeletal muscles (9,10). The disruption of the dystrophin-sarcoglycan complex causes increased calcium influx and chronic elevation of intracellular calcium, resulting in cell damage in the δ-sarcoglycan-deficient hamsters (11). Abnormality in dystrophin is observed in human dilated or ischemic cardiomyopathic hearts (12).
In this study, we examined whether chronic inhibition of ASK1 activation by transcoronary gene transfer using recombinant adeno-associated virus (rAAV) (13) could attenuate progression of cardiac remodeling in TO-2 cardiomyopathic hamsters. ASK1 inhibition was remarkably effective for preventing progression of cardiac remodeling and heart failure. Thus, suppression of ASK1 may constitute a novel therapeutic strategy for the treatment of patients with heart failure.
Male F1B (normal hamster) and TO-2 hamster strains were obtained from Bio Breeders (Watertown, Massachusetts). All animal protocols were approved by the Yamaguchi University School of Medicine Animal Subject Committee.
Recombinant adeno-associated virus type-2 was generated by 3 plasmid cotransfection method. ASKΔN(KR) was a 5′-deletion mutant of ASK(KR) starting at 1,945 bp downstream of the translation initiation site. Recombinant adenovirus vectors were constructed using ViraPower Adenoviral Expression System (Invitrogen, Carlsbad, California).
Animal procedures and hemodynamic measurement
Recombinant adeno-associated virus was delivered into the coronary arteries of 10-week-old TO-2 hamsters following a previously described protocol (13). A total of 1 to 4 × 1011viral particles per 100 g of body weight was injected through the transcoronary route. Hamsters were anesthetized with sodium pentobarbital (50 mg/kg intraperitoneally), followed by echocardiographic analysis with an HDI-5000 ultrasound machine (Philips, Eindhoven, the Netherlands) equipped with a 15-MHz linear probe. Hamsters were underwent LV pressure measurement as previously described (13).
Evaluation of apoptosis
Triple staining with the terminal deoxynucleotidyl transferase biotin-dUTP nick end labeling (TUNEL) assay was performed using an in situ apoptosis detection kit (Takara, Otsu, Japan). The cells undergoing apoptosis were labeled with fluorescein-dUTP and observed under a confocal fluorescence microscope. Immunohistochemical staining using anti-activated caspase 3 antibody (Abcam Inc., Cambridge, Massachusetts) was performed with paraffin-embedded sections.
Immune complex kinase assay and Western blots
The activity of ASK1 or MKK6 was measured with an immune complex kinase assay as previously described (4). Protein homogenates (60 μg/lane) were subjected to Western blot analysis using the antibodies against JNK, p38 and extracellular signal-regulated kinase (ERK) (Santa Cruz Biotechnology, Santa Cruz, California), phospho-p38, phospho-JNK, and phosho-ERK (Cell Signaling Technology, Beverly, Massachusetts), α-, β-, and γ-sarcoglycans, and dystrophin (rod domain; Novocastra, Newcastle, United Kingdom).
Measurement of calpain activity and intracellular calcium level
Rat neonatal ventricular cardiomyocytes were prepared as previously reported (4) and recombinant adenovirus vectors expressing ASKΔN(KR) (AdV/ASKΔN(KR)) or LacZ (AdV/LacZ) were infected at a multiplicity of infection of 25. Twenty-four hours after adenovirus infection, intracellular calpain proteolytic activity was estimated by t-butoxycarbonyl-leucyl-methionine-7-amino-4-chloromethylcoumarin (Boc-Leu-Met-CMAC) (Molecular Probes, Eugene, Oregon) as we previously reported (14) using an excitation wavelength of 340 nm and an emission wavelength of 430 nm. Calcium overload was induced by changing an incubation solution to a sodium-free one. We measured at least 3 cardiomyocytes in each experiment. To measure intracellular calcium levels, 4 μmol/l fura-2 AM (Molecular Probes) was loaded into cardiomyocytes for 10 min. The fluorescent light emitted at 510 nm and excited at 340 and 380 nm after the change to the sodium-free solution was measured, and R340/380was calculated.
Results are shown as mean ± SEM. Statistical analyses were performed by using Statcel 2 software (OMS publishing Inc., Tokorozawa, Japan). Comparisons between 2 groups were performed with the Student ttest. A 1-way analysis of variance (ANOVA) with the Bonferroni post-hoc test was used for multiple comparisons. A probability value of <0.05 was considered statistically significant.
Generation of rAAV expressing a dominant negative form of ASK1
First, we attempted to generate rAAV expressing a dominant-negative form of ASK1. Although ASK(KR), with a critical Lys residue in the kinase domain replaced by Arg, is a dominant-negative mutant of ASK1 (3), it is too large in size to be incorporated into rAAV. Thus, we deleted the amino-terminal region of ASK(KR) (Met1to Glu648) to generate ASKΔN(KR) (Fig. 1A).Infection of cardiomyocytes with AdV/ASKΔN(KR) did not result in MKK6 activation (Fig. 1B), suggesting that ASKΔN(KR) is kinase inactive. Phenylephrine-induced MKK6 activation was attenuated in AdV/ASKΔN(KR)-infected cardiomyocytes in a dose-dependent manner, indicating that ASKΔN(KR) can function as a dominant-negative form.
Myocardial gene delivery of rAAV expressing ASKΔN(KR) (rAAV/ASKΔN(KR))
A total of 44 TO-2 cardiomyopathic hamsters were evaluated by means of echocardiography at 10 weeks of age and subjected to a transcoronary in vivo gene transfer. Four hamsters died during the gene transfer procedure, and the remaining 40 cardiomyopathic hamsters were randomly assigned to 3 groups (13 cardiomyopathic hamsters received rAAV/LacZ treatment, 19 received rAAV/ASKΔN(KR), and 8 received only saline without virus). These 3 groups did not show any statistical differences in cardiac function before gene transfer. One hamster treated with rAAV/LacZ died as the result of congestive heart failure 35 days after gene transfer and was excluded from analysis thereafter. The overall transfection efficiency of rAAV/LacZ at 14 weeks after gene transfer was approximately 50% as assessed by LacZ staining of the LV (Fig. 1C), which was similar to that in our previous reports (13), and no rAAV/LacZ expression was observed in liver. We detected a high expression level of HA-tagged ASKΔN(KR) in rAAV/ASKΔN(KR)-infected heart, whereas the endogenous ASK1 expression level was too low to be detected under the conditions used (Fig. 1D). ASK1 activity in LacZ-infected hamster hearts increased, which was significantly attenuated in ASKΔN(KR)-infected hearts 4 weeks after gene transfer (Fig. 1E).
Echocardiographic analysis indicated a reduction in LV fractional shortening, LV ejection fraction, diastolic wall thickness of interventricular septum or LV posterior wall (LVPWT), and an increase in LV end-diastolic dimension or LV end-systolic dimension already were evident at 10 weeks of age, when the in vivo gene transfer was performed (p < 0.001 vs. normal hamsters, by 1-way ANOVA with the Bonferroni post hoc test, 3 comparisons) (Figs. 2Bto 2G). In the rAAV/ASKΔN(KR) group, LV fractional shortening, LV ejection fraction, diastolic wall thickness of interventricular septum, or LVPWT was significantly greater and LV end-diastolic dimension or LV end-systolic dimension was significantly lower than in rAAV/LacZ group both 6 and 14 weeks after gene transfer. The index of LV diastolic wall stress, LVDd/LVPWT, was also significantly lower in rAAV/ASKΔN(KR) than in rAAV/LacZ group (Fig. 2H).
The rAAV/LacZ and saline groups did not show any differences (data not shown), excluding any artificial and independent effects produced by the rAAV vectors. The rAAV/ASKΔN(KR) group showed a significantly higher maximum first derivative of LV pressure and significantly lower minimum first derivative of LV pressure compared with rAAV/LacZ group 14 weeks after gene transfer (Fig. 3).The mRNA levels of atrial natriuretic factor, brain natriuretic peptide, and β-myosin heavy chain were significantly attenuated in rAAV/ASKΔN(KR) compared with those in rAAV/LacZ group (p = 0.014 for atrial natriuretic factor, 0.003 for brain natriuretic peptide, and 0.015 for β-myosin heavy chain, 1-way ANOVA with the Bonferroni post hoc test, 3 comparisons) (Fig. 4).These findings indicate that the inhibition of ASK1 suppressed functional, structural, and biochemical features of the progression of cardiac remodeling.
Reduction in histological abnormalities by rAAV/ASKΔN(KR)
The TO-2 hamster shows histological abnormalities in the heart, including an increase in calcified lesion and interstitial fibrosis (15). The rAAV/ASKΔN(KR)-infection reduced calcification and fibrosis (Figs. 5Ato 5F). In addition, the expression level of type III collagen mRNA had significantly decreased in rAAV/ASKΔN(KR) compared with that in rAAV/LacZ group (p = 0.033 by 1-way ANOVA with the Bonferroni post hoc test, 3 comparisons) (Fig. 5G).
Reduction in apoptotic cardiomyocyte death by rAAV/ASKΔN(KR)
The in situ TUNEL assay 14 weeks after gene transfer revealed that TUNEL-positive cells contained condensed chromatin and were identified as cardiomyocytes by anti-α-sarcomeric actin staining (Fig. 6A).The number of TUNEL-positive cardiomyocytes had significantly decreased in rAAV/ASKΔN(KR) compared with that in rAAV/LacZ group (Fig. 6B). Furthermore, the number of activated caspase 3-positive cardiomyocytes had significantly decreased in rAAV/ASKΔN(KR) group (Figs. 6C and 6D).
The effects of rAAV/ASKΔN(KR) on downstream MAPK cascade 4 weeks after gene transfer
The phosphorylation levels of p46 or p54-JNK had significantly decreased in rAAV/ASKΔN(KR) compared with those in rAAV/LacZ group (Figs. 7Aand 7B), whereas those of p38 in rAAV/ASKΔN(KR) and rAAV/LacZ groups were not significantly different (Figs. 7A and 7C). The ERK phosphorylation level showed no differences between both groups (Figs. 7A and 7D).
Attenuation of dystrophin cleavage and calpain activity by gene therapy
Cardiomyopathic hamsters exhibit proteolysis of myocardial dystrophin (10). We could not detect any expression of α-, β-, or γ-sarcoglycans in either rAAV/LacZ or rAAV/ASKΔN(KR) group (Fig. 8A).Protein level of full-length dystrophin had decreased and that of degraded dystrophin had increased in rAAV/LacZ group compared with those in normal hamsters (Fig. 8B). However, the proteolysis of dystrophin was markedly suppressed in rAAV/ASKΔN(KR) compared with that in rAAV/LacZ group (Fig. 8B). Considering that intracellular calcium in cardiomyocytes of this hamster strain is elevated (11) and that dystrophin is a well-defined substrate of calpain, a calcium-dependent protease (16), it is possible that suppression of dystrophin degradation may be mediated through inhibition of calpain activity (10). Western blot analysis indicated that the autolyzed form of μ-calpain, which is an activated form of calpain (17), had clearly increased in rAAV/LacZ group compared with that in normal hamsters, and it had markedly decreased in rAAV/ASKΔN(KR) compared with that in rAAV/LacZ group (Fig. 8C).
Next, we examined whether ASK1 inhibition could attenuate calpain activation in isolated cardiomyocytes. Intracellular calcium levels were nearly identical for AdV/LacZ-infected and AdV/ASKΔN(KR)-infected cardiomyocytes throughout experiment (Fig. 8D). ASK1 was activated as early as 10 min after Ca2+overload (Fig. 8E). The calcium-induced calpain activity, expressed as the ratio of increase in fluorescence intensity, was significantly lower in AdV/ASKΔN(KR)-infected than in AdV/LacZ-infected cardiomyocytes (p = 0.031 by 1-way ANOVA with the Bonferroni post-hoc test, 3 comparisons) (Fig. 8F).
In the present study, we demonstrated that in vivo gene transfer of a dominant-negative mutant of ASK1 via rAAV vector efficiently prevented heart failure progression in cardiomyopathic hamsters for 3 months. ASK1 inhibition in hamster cardiomyocytes not only prevented chamber dilation but also preserved LV systolic and diastolic function. Moreover, long-term expression of the dominant-negative mutant of ASK1 significantly inhibited cardiac interstitial fibrosis. This is the first demonstration that ASK1 inhibition is beneficial for preventing heart failure progression even after the onset of hereditary cardiomyopathy.
Apoptosis has been identified as an essential process in the progression of heart failure (18). ASK1 activation induces apoptosis in cardiomyocytes, whereas oxidative stress-induced apoptosis is suppressed in ASK1-/-cardiomyocytes (5). In addition, the prevention of LV remodeling was accompanied by a reduction in the appearance of apoptotic cardiomyocytes in ASK1-/-hearts (5). In this study, we detected reduced apoptosis in ASKΔN(KR)-infected hearts. The ASK1-JNK pathway has been reported to be a crucial element of apoptosis (5). Recently, it was reported that β-agonist-mediated cardiomyocyte apoptosis was attenuated by heat shock protein Hsp20 through the inhibition of ASK1-JNK/p38 pathway (19). We observed that rAAV/ASKΔN(KR) selectively inhibited JNK activation. These results suggest that the attenuated caspase-dependent cardiomyocyte apoptosis, mediated through an inhibition of ASK1-JNK activation by rAAV/ASKΔN(KR), will be a mechanism of suppression of heart failure progression in cardiomyopathic hamsters.
Interestingly, the cleavage of dystrophin with the disease progression also was markedly prevented by gene transfer. Dystrophin-glycoprotein complex provides a strong mechanical link from the intracellular cytoskeleton to the extracellular matrix. Cardiomyocyte degeneration has been attributed to membrane defects as a greater fragility toward mechanical stress (9) or increased permeability to Ca2+(11). The resultant increase in intracellular Ca2+may lead to the activation of the Ca2+-dependent protease calpain as observed in an mdx mice model lacking dystrophin (20). Calpain in turn will degrade dystrophin and various myofibrillar proteins (21), ultimately resulting in necrosis and apoptosis (22). Moreover, degradation of dystrophin will impair sarcolemma integrity to enter a vicious cycle to advance heart failure (10). Our study showed that Ca2+overload activated both of ASK1 and calpain. Inhibition of ASK1 by the dominant-negative mutant completely abrogated calpain activation, whereas it did not affect the increase in intracellular Ca2+. These data suggest that ASK1 activation is the upstream event of calpain activation in cardiomyocytes. Thus, it is possible that calpain-mediated dystrophin cleavage was abrogated by the inhibition of ASK1. Precise molecular mechanism that explains how ASK1 is involved in the calpain activation remains to be elucidated.
The disruption of dystrophin is not a unique feature of δ-sarcoglycan-deficient cardiomyopathic hamster. Both acute and chronic heart failure models in rats time-dependently caused dystrophin disruption in myocardium (15). Dystrophin is disrupted in hearts of patients with dilated or ischemic end-stage cardiomyopathy (12). These results suggest that disruption of myocardial dystrophin appears to be a common pathway to advanced heart failure. With the observation that ASK1 ablation inhibited the progression of cardiac remodeling induced by pressure overload and after myocardial infarction, the therapeutic strategy targeting ASK1, thus, will not be limited to δ-sarcoglycan-deficient cardiomyopathy but will be applied to multiple acquired and genetic forms of heart failure. However, further investigation using other experimental models such as myocardial infarction or larger animals will be necessary to conclude the efficacy of ASK1 inhibition in the treatment of heart failure.
Long-term suppression of ASK1 activation with rAAV/ASKΔN(KR) gene transfer ameliorated heart failure progression by preventing cardiomyocyte apoptosis and by inhibiting calpain-mediated dystrophin cleavage in the cardiomyopathic hamsters. ASK1 could thus be a novel promising therapeutic target not only to apply viral vector-mediated gene therapy but also to develop a novel drug for the treatment of heart failure.
↵1 Drs. Hikoso and Yamaguchi held a postdoctoral fellowship for Center of Excellence Research from the Ministry of Education, Culture, Sports, Science and Technology and from Japan Society for the Promotion of Science, respectively.
↵2 Drs. Hikoso and Ikeda contributed equally to this work.
This work was supported by a Grant-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology, Japan (16590683) and research grants from Hyogo Science and Technology, Takeda Science Foundation, and Sumitomo Foundation to Dr. Otsu.
- Abbreviations and Acronyms
- analysis of variance
- apoptosis signal-regulating kinase 1
- extracellular signal-regulated kinase
- c-Jun NH2-terminal kinase
- left ventricular
- left ventricular posterior wall
- mitogen-activated protein
- mitogen-activated protein kinase kinase
- recombinant adeno-associated virus
- terminal deoxynucleotidyl transferase biotin-dUTP nick end labeling
- Received October 24, 2006.
- Revision received February 28, 2007.
- Accepted March 20, 2007.
- American College of Cardiology Foundation
- Hayashidani S.,
- Tsutui H.,
- Shiomi T.,
- et al.
- Ichijo H.,
- Nishida E.,
- Irie K.,
- et al.
- Hirotani S.,
- Otsu K.,
- Nishida K.,
- et al.
- Yamaguchi O.,
- Higuchi Y.,
- Hirotani S.,
- et al.
- Sakamoto A.,
- Ono K.,
- Abe M.,
- et al.
- Ikeda Y.,
- Martone M.,
- Gu Y.,
- et al.
- Toyo-Oka T.,
- Kawada T.,
- Nakata J.,
- et al.
- Iwata Y.,
- Katanosaka Y.,
- Arai Y.,
- Komamura K.,
- Miyatake K.,
- Shigekawa M.
- Vatta M.,
- Stetson S.J.,
- Jimenez S.,
- et al.
- Goll D.E.,
- Thompson V.F.,
- Li H.,
- Wei W.,
- Cong J.
- Fan G.C.,
- Yuan Q.,
- Song G.,
- et al.
- Spencer M.J.,
- Croall D.E.,
- Tidball J.G.
- Gil-Parrado S.,
- Fernandez-Montalvan A.,
- Assfalg-Machleidt I.,
- et al.