Oxidative Stress Causes Heart Failure with Impaired Mitochondrial Respiration*

Elderly people insidiously manifest the symptoms of heart failure, such as dyspnea and/or physical disabilities in an age-dependent manner. Although previous studies suggested that oxidative stress plays a pathological role in the development of heart failure, no direct evidence has been documented so far. In order to investigate the pathological significance of oxidative stress in the heart, we generated heart/muscle-specific manganese superoxide dismutase-deficient mice. The mutant mice developed progressive congestive heart failure with specific molecular defects in mitochondrial respiration. In this paper, we showed for the first time that the oxidative stress caused specific morphological changes of mitochondria, excess formation of superoxide (\batchmode \documentclass[fleqn,10pt,legalpaper]{article} \usepackage{amssymb} \usepackage{amsfonts} \usepackage{amsmath} \pagestyle{empty} \begin{document} \(\mathrm{O}_{2}^{{\bar{{\cdot}}}}\) \end{document}), reduction of ATP, and transcriptional alterations of genes associated with heart failure in respect to cardiac contractility. Accordingly, administration of a superoxide dismutase mimetic significantly ameliorated the symptoms. These results implied that \batchmode \documentclass[fleqn,10pt,legalpaper]{article} \usepackage{amssymb} \usepackage{amsfonts} \usepackage{amsmath} \pagestyle{empty} \begin{document} \(\mathrm{O}_{2}^{{\bar{{\cdot}}}}\) \end{document} generated in mitochondria played a pivotal role in the development and progression of heart failure. We here present a bona fide model for human cardiac failure with oxidative stress valuable for therapeutic interventions.

energy consumption, a phenomenon that is best described as mechanoenergetic uncoupling. Despite the markedly impaired work of the left ventricle in this process, the oxygen used by muscle contraction remains relatively unchanged, which results in a decline in the contractile function of the myocardium.
The mechanisms responsible for the progression of heart failure are emerging, whereas there is increasing evidence to suggest that reactive oxygen species (ROS) play a major pathological role (2,3). In fact, several studies on human cases have revealed increased ROS production in the hearts of individuals with congestive heart failure. Biomarkers for ROS have been detected in the pericardial fluid (4) as well as in the peripheral blood (5, 6) of heart failure patients. Although it is still controversial whether antioxidative enzymes are inactivated in failing hearts (7,8), polymorphisms in the candidate genes that are responsive to oxidative stress affect the susceptibility to and contribute to the risk of heart failure (2). It is therefore hypothesized that heart failure is attributable to an imbalance between the generation of ROS and cellular antioxidative defensive mechanisms.
Superoxide (O 2 . ), one of the ROS generated in mitochondrial respiration, is involved in a variety of biological processes in the myocardium, including alterations in mitochondrial DNA (9), contractile proteins (10), ion channels (11), or mitochondrial respiratory enzymes (12,13). Manganese superoxide dismutase (Mn-SOD) is a principal scavenger enzyme located in mitochondrial matrix, which detoxifies O 2 . by dismutation, protecting cells from oxidative stress. Interestingly, polymorphisms in the targeting signal of the Mn-SOD gene affect the efficiency of its transport to mitochondria. The allele frequency is then associated with idiopathic dilated cardiomyopathy (14,15). These genetic data suggested that Mn-SOD plays an important role in the susceptibility to and/or the progression of dilated cardiomyopathy. In previous studies, two groups independently reported on Mn-SOD-deficient mice (16,17). The Mn-SOD deficiency on a CD1 background resulted in neonatal death by day 10 from severe dilated cardiomyopathy, liver dysfunction, and metabolic acidosis (16). Other Mn-SOD-deficient mice in a mixed genetic background of C57BL/6 and 129/Sv, on the other hand, died by day 18 with neuronal degeneration, particularly in the basal ganglia and brainstem (17). In this context, it was hard to investigate the pathological consequences of oxidative damages in adult tissues by using these Mn-SOD-deficient mice. These authors also argued that the phenotypes are too complex to sequestrate the specific aging processes in each tissue of mutant mice in vivo. In the present study, we successfully generated heart/muscle-specific Mn-SOD-deficient mice (H/M-Sod2 Ϫ/Ϫ ) using the Cre-loxP system under the control of the muscle creatine kinase (MCK) promoter and established a murine model for congestive heart failure as well as age-related physical disability induced by oxidative stress.
Since heart failure still remains a significant health problem, with mortality rates as high as 45% within 1 year (1), more effective treatments should be established. Since the biological effects of oxidative stress in heart failure have been precisely analyzed in the last decade, we expect an emergence of new cardiovascular medicine or new nutritional interventions that will improve the quality of life in patients with heart failure. Many of the antioxidants and oxidative stress-related medicines currently under investigation are controversial (3,18). With this model, we provide a means to evaluate the efficacy of medicine or interventions over a physiological as well as pathological range of heart functions and failures.

EXPERIMENTAL PROCEDURES
Animals and Genotyping-The generation of Mn-SOD flox mice was described previously (19). The Mn-SOD flox mice were back-crossed to C57BL6/CrSlc mice for five or six generations. The cross-breeding of homozygous Mn-SOD flox mice with MCK-Cre transgenic mice of FVB background (20) gave rise to H/M-Sod2 Ϫ/Ϫ mice. All genotyping of the Cre transgene and the Mn-SOD flox was performed by PCR using genomic DNA isolated from the tail tip. The primers for identifying carriers of the MCK-Cre transgene (5Ј-TTC CAG CTA GAG AGA CTG AGT CCC-3Ј and 5Ј-TCG ACC AGT TTA GTT ACC C-3Ј) and Mn-SOD flox allele (5Ј-TTA GGG CTC AGG TTT GTC CAG AA-3Ј, 5Ј-CGA GGG GCA TCT AGT GGA GAA-3Ј, and 5Ј-AGC TTG GCT GGA CGT AA-3Ј) were used. The deleted alleles were confirmed as described previously (19). DNA isolated from cytomegalovirus enhancer/chicken ␤-actin (CAG) promoter Cre Mn-SOD-deficient neonates was used as a positive control (19). Mice were maintained and studied according to protocols approved by the Animal Care Committee of the Tokyo Metropolitan Institute of Gerontology.
Western Blot Analysis-Western blot analysis was performed with heart, skeletal muscle (tibialis anterior), brain, liver, and kidney homogenates (5 g each) as previously described (19 Generation of an Antibody against Mn-SOD for Immunohistochemistry-An immunogen peptide corresponding to residues Lys 25 -Ala 37 (KHSLPDLPYDYGA) of mouse Mn-SOD with an additional cysteine residue plus amide at the C terminus was synthesized. The synthetic peptide corresponding to the N-terminal region of the mature form of mouse Mn-SOD (21) was immunized to rabbits as described previously (22). The specific antibody was affinity-purified using the immunogen peptide immobilized on FMP-activated Cellulofine (Seikagaku Kogyo, Tokyo, Japan).
Histological and Histochemical Studies-Organs were dissected, fixed overnight in a 20% formalin neutral buffer solution (Wako, Osaka, Japan), embedded in paraffin, and sectioned on a microtome using standard techniques. To detect the fibrosis or glycogen accumulation in tissues, sections were stained by Azan staining or periodic acid-Schiff (PAS) staining methods, respectively. For immunohistochemistry, 4-m deparaffinized transverse sections prepared from heart or tibialis anterior were immunostained with an anti-Mn-SOD antibody (1:1000) using a Vectastain elite ABC kit (Vector Laboratories, Burlingame, CA) according to the manufacturer's protocol. For enzymatic histochemical staining, tissues were frozen in isopentane in liquid nitrogen and embedded in Tissue Tek O.C.T. compound on dry ice. Sections were cut 8 m thick and mounted on silane-coated slide glasses. Frozen sections were dried and incubated in 50 mM sodium phosphate (pH 7.4), 84 mM succinate acid, 0.2 mM phenazine methasulfate, 2 mg/ml nitro blue tetrazolium, and 4.5 mM EDTA for SDH (Complex II) activity or in 50 mM sodium phosphate (pH 7.4), 1.0 mg/ml 3,3Ј-diaminobenzidine, 24 units/ml catalase (Wako), 1 mg/ml cytochrome c (Wako), and 75 mg/ml sucrose for COX (Complex IV) activity. These reactions were performed in the dark at room temperature for 20 min. The nuclei with DNA fragmentation on 4-m deparaffinized sections were stained with an ApopTag Plus fluorescein in situ apoptosis detection kit (Serologicals Corp., Norcross, GA).
Echocardiography-The mice were anesthetized with 2.5% avertin (8 l/g) and echocardiography was performed using ultrasonography (SONOS-5500, equipped with 15-MHz linear transducer; Philips Medical Systems) (23). The heart was imaged in the two-dimensional parasternal short axis view, and an M-mode echocardiogram of the midventricle was recorded at the level of the papillary muscles. Heart rate, anterior and posterior wall thickness, and end diastolic and end systolic internal dimensions of the left ventricle were obtained from the M-mode image. Activities of Mitochondrial Respiratory Chain Enzymes and ATP Synthesis-Mitochondria were isolated from heart, tibialis anterior muscle, and liver by homogenization, followed by differential centrifugation (24). Oxidative phosphorylation (OXPHOS) enzyme activity was normalized for the protein concentration. Succinate-cytochrome c reductase (Complex II ϩ III) activity and NADH-cytochrome c reductase (Complex I ϩ III) activity were measured as described previously (24). ATP synthesis was measured in intact heart and liver mitochondria. Incubation buffer (250 mM sucrose, 5 mM KCl, 5 and P2 primers as indicated in A. A 401-bp DNA fragment amplified from the deletion allele was detected in heart (H) and skeletal muscle (Sm) but not in brain (B), lung (Lu), liver (Li), and kidney (K ). DNA from CAG-Cre Mn-SOD-deficient neonates was used as a positive control. C, Western blot analysis of Mn-SOD protein in 16-week-old H/M-Sod2 Ϫ/Ϫ and control mice. Protein extracts from heart (H), skeletal muscle (Sm), brain (B), Liver (Li), or kidney (K) of H/M-Sod2 Ϫ/Ϫ or control mice were immunoblotted with anti-Mn-SOD or anti-glyceraldehyde-3-phosphate dehydrogenase antibodies. D, immunohistochemical staining of heart and tibialis anterior muscles from 12-week-old mice of the indicated genotypes using anti-Mn-SOD antibody (scale bar, 10 m). mM KH 2 PO 4 , 5 mM MgCl 2 , 1 mM EDTA, and 10 mM Tris-HCl, pH 7.4) was added to test tubes followed by the addition of mitochondria (0.01 mg of protein/ml). The tubes were incubated at 25°C for 10 min. After that, 0.5 mM ADP, 5 mM malate, and 0.5 mM glutamate as substrates were added to test tubes and incubated at 25°C for 5 min. Then the test tubes were placed on ice and centrifuged at 10,000 ϫ g, 4°C, for 10 min. A 190-l aliquot of the supernatant was added to 10 l of luciferase and luciferin (Wako), and bioluminescence was assessed on a Lumat LB9507 luminometer (Berthold Technologies, Bad Wildbad, Germany). Results were expressed as nmol of ATP/mg of protein. ATP content was measured with a cell Titer-Glo luminescent cell viability assay (Promega Corp., Madison, WI) according to the manufacturer's protocol.

Transmission Electron Micrographs of Myocardium and
RNA Preparation, Affymetrix Microarray, and RT-PCR-Total RNA was extracted from hearts of three 16-week-old mice with an RNeasy Fibrous Tissue Mini Kit (Qiagen, Valencia, CA). For the microarray analysis, the synthesis of cDNA and array hybridization was conducted with Affymetrix MOE430 2.0 array (Affymetrix, Santa Clara, CA) according to the manufacturer's protocol. The raw microarray data files are available on the World Wide Web at www.ncbi.nlm.nih.gov/geo/ (Gene Expression Omnibus accession number GSE2236).
Rescue Experiments-Manganese 5,10,15,20-tetrakis (4-benzoic acid) porphyrin (MnTBAP) (Calbiochem) dissolved in PBS at 1 mg/ml was injected intraperitoneally into mutant mice once a day as previously described (27). The injections, the dose of which was 10 g/g of body weight/day, began from 8 weeks of age for 4 -8 weeks. PBS was injected into the littermate mutant mice as control.
Analysis of Physical and Cardiac Contractile Activities-Mice were voluntarily exercised using a running wheel apparatus (KI-101; Aptec, Kyoto, Japan) as described previously (28). The rotarod task was analyzed using a rotarod apparatus (MK-630; Muromachi Kikai, Tokyo, Japan). The period for which a mouse could remain on a rotating axis (30-mm diameter; rota-tional speed 20 rpm) without falling was measured. The test was stopped after an arbitrary limit of 300 s, and three sessions of training were required before mice attained a stable performance. Isolated hearts from 12-week-old mice were excised quickly to establish Langendorff perfusion and perfused with modified Krebs-Henseleit solution (containing 116.0 mM NaCl, 25.0 mM NaHCO 3 , 2.5 mM CaCl 2 , 1.2 mM MgSO 4 , 4.7 mM KCl, 1.2 mM KH 2 PO 4 , and 5.5 mM glucose, pH 7.4) in a retrograde direction at a constant flow rate of 2.0 ml/min without recirculation. The perfusate was warmed to 38°C and oxygenated with a 95% O 2 , 5% CO 2 gas mixture to elevate the PO 2 to over 400 mm Hg. A hook through the apex of the heart was connected to an isometric force transducer for monitoring tension and heart rate. The initial loaded tension was adjusted to the same force in each experiment. The cardiac contractile activity was evaluated with the tension rate product (tension times heart rate). The data were normalized with heart weight.
Statistical Analysis-We analyzed data using Student's unpaired t test and considered p values of Ͻ0.05 to be statistically significant. Data are expressed as the mean Ϯ S.D.

Generation of Heart/Muscle-specific Mn-SOD-deficient
Mice-In order to investigate the physiological as well as pathological role of Mn-SOD in the heart and muscle, we generated conditional Mn-SOD knock-out mice using the Cre-loxP system. We used MCK-Cre transgenic mice for the selective expression of Cre protein in muscle tissues (20). As shown in Fig. 1, cross-breeding of homozygous Mn-SOD flox mice (control) with MCK-Cre transgenic mice gave rise to H/M-Sod2 Ϫ/Ϫ . Genomic DNAs extracted from various tissues were analyzed by PCR for the detection of the deleted fragment from the genomic Mn-SOD gene. A 401-bp DNA fragment corresponding to the deletion allele was specifically amplified by PCR from the heart and skeletal muscle of H/M-Sod2 Ϫ/Ϫ mice, whereas no fragment was amplified in other tissues of H/M-Sod2 Ϫ/Ϫ mice or in any tissues of control mice (Fig. 1B). Western blot analyses further showed a specific loss of Mn-SOD expression in the heart and skeletal muscle of H/M-Sod2 Ϫ/Ϫ mice (Fig. 1C), but no loss was observed in control mice. In an immunohistochemical analysis, specific immunoreactivity for Mn-SOD was detected in the endothelial cells of the heart or muscle but not in the cardiac myocytes or the striated muscle cells of the tibialis anterior (TA) from H/M-Sod2 Ϫ/Ϫ mice. On the other hand, specific immunoreactivity for Mn-SOD was detected in the myocardium and in the striated muscle from the TA of control mice (Fig. 1D).
Heart/Muscle-specific Mn-SOD-deficient Mice Developed Dilated Cardiomyopathy Associated with Progressive Physical Disabilities-In the neonatal stage, we were unable to find any differences in the outer appearance or body size between H/M-Sod2 Ϫ/Ϫ and control mice. However, at 8 weeks of age, the H/M-Sod2 Ϫ/Ϫ mice began to exhibit growth retardation. At 16 weeks of age, the H/M-Sod2 Ϫ/Ϫ mice showed a 25% reduction in body weight compared with control mice without distinct muscle atrophy (Fig. 2, A and  C). Phenotypically, H/M-Sod2 Ϫ/Ϫ mice developed signs of fatigue at as early as 8 weeks of age, when some of these animals began to die. By 22 weeks of age, all H/M-Sod2 Ϫ/Ϫ mutant mice died, with a median survival rate of 15.4 Ϯ 4.0 weeks (Fig. 2B). When examined macroscopically, all of the hearts of H/M-Sod2 Ϫ/Ϫ mice showed cardiac enlargement at 16 weeks of age without an exception ( Fig. 2A). Under normal diet, the H/M-Sod2 Ϫ/Ϫ mice exhibited 51% reduction in food intake compared with control mice (Fig. 2D). In order to evaluate daily physical activities of the H/M-Sod2 Ϫ/Ϫ mutant mice, we placed a running wheel apparatus in their cages. The H/M-Sod2 Ϫ/Ϫ mutant mice did not exhibit any signs of ataxia but hardly ran on the apparatus, whereas control mice ran more than 10 km every day after 4 days of the exercise session (Fig. 2E). When heart weight was standardized to body weight, the hearts of H/M-Sod2 Ϫ/Ϫ mice were found to be 2.1-fold heavier at 2 months of age and 2.7-fold heavier at 4 months of age than those of control mice (Fig.  2F, Table 1).
To determine whether deficiency of Mn-SOD in heart and muscle would affect the cardiac function, cardiac performance was evaluated by means of echocardiography in 2-month-old mice (Fig. 3, Table 1). Compared with control mice, cardiac contractility was significantly depressed in H/M-Sod2 Ϫ/Ϫ mice as assessed by fractional shortening (FS) and ejection fraction (EF) (Fig. 3B, Table 1). The LV enddiastolic and end-systolic diameters were significantly increased in H/M-Sod2 Ϫ/Ϫ mice compared with control mice (Fig. 3B, Table 1). The decreased contractility and LV dilatation were also observed at 4 months compared with control mice. However, diastolic intraventricular septum thickness (IVSd), diastolic LV posterior wall thickness (LVPWd), and systolic LV posterior wall thickness (LVPWs) were not significantly increased in H/M-Sod2 Ϫ/Ϫ mice compared with control mice (Fig. 3, Table 1).
The Myocardium Showed Specific Pathological Findings Compatible with Typical Idiopathic Dilated Cardiomyopathy-Transverse sections of hearts from the H/M-Sod2 Ϫ/Ϫ mice showed a marked dilatation of both left and right ventricles, which is compatible with the end stage of dilated cardiomyopathy (Fig. 4, A and  B). High power photomicrographs of the LV wall showed myocardial degeneration, myocyte disarray, vacuolization, and bizarre myocardial cells with irregular myofilaments and pleomorphic nuclei (Fig. 4C). In the histological sections with Azan staining, diffuse fibrotic scars surrounded myocardial cells. Some of the thickened fibrotic foci were due to necrotic changes of the myocardium, whereas the majority of the thin layer of interstitial fibrosis was organized through the longstanding dilatation of ventricles (Fig. 4, E and F). In the sections with PAS, PAS-positive substances, such as glycogen and mucin, were abnormally distributed in the interstitial space (Fig. 4, G and H). Electron micrographs of the LV wall of H/M-Sod2 Ϫ/Ϫ mutant mice showed small mitochondria associated with scattered abnormal vacuoles (Fig. 5A). However, we were unable to find any ultrastructural changes in sarcomeric structures between H/M-Sod2 Ϫ/Ϫ mice and control mice. The cristae of mutant mitochondria were rough, irregular, abnormally   Views of fluorescein staining using anti-digoxigenin conjugate (green) were merged with those of 4Ј,6-diamidino-2-phenylindole staining (blue). E, Western blot analysis of protein extracted from heart (H), skeletal muscle (Sm), and liver (L) of 15-week-old H/M-Sod2 Ϫ/Ϫ mice using antibodies against activated caspase-3, glyceraldehyde-3-phosphate dehydrogenase, and Mn-SOD. We used NIH-3T3 cells treated with staurosporine as apoptotic positive control (P). F, ATP content in heart (n ϭ 9), TA muscle (n ϭ 6), and liver (n ϭ 6) of 15-week-old H/M-Sod2 Ϫ/Ϫ mice (*, p Ͻ 0.001).
wound, and concentrated in the central zone of the matrix (Fig.  5, C and D). Similar but modest ultrastructural findings were observed in the skeletal muscle of the tibialis anterior (data not shown). However, we found no abnormal crystals or droplets in mitochondria as shown in the case of human dilated cardiomyopathy (29).

The Heart/Muscle-specific Mn-SOD-deficient Mice Failed to Show Apoptotic Cell Death in Myocardium and Skeletal
Muscle-To determine the manner of cell death in myocardium and skeletal muscle, we investigated the nuclear morphology by electron microscopy, TUNEL staining, expression of cleaved caspase-3, and ATP content in 15-week-old H/M-Sod2 Ϫ/Ϫ mice and control mice. In the study of electron microscopy as far as we investigated, we hardly detected the apoptotic nuclei with DNA fragmentation in myocardium of H/M-Sod2 Ϫ/Ϫ mice as well as control mice (Fig. 6, A and B). Furthermore, no difference was observed in the number of TUNELpositive apoptotic cells between H/M-Sod2 Ϫ/Ϫ mice and control mice (Fig. 6, C and D). We also failed to detect the activated form of caspase-3 in myocardium and skeletal muscle of H/M-Sod2 Ϫ/Ϫ mice and control mice (Fig. 6E). Moreover, we examined ATP content in heart, skeletal muscle, and liver, because ATP is required to induce the activation of caspase-3. ATP contents in heart and skeletal muscle of H/M-Sod2 Ϫ/Ϫ mice both decreased to about 30% of those of control mice (Fig. 6F), whereas we failed to detect any difference in livers between H/M-Sod2 Ϫ/Ϫ mice and control mice (Fig. 6F).

The Heart/Muscle-specific Mn-SOD-deficient Mice Showed Suppressed OXPHOS in Myocardium-
To understand the biochemical alterations in pathogenesis of cardiomyopathy, we examined the mitochondrial respiratory functions in the heart of H/M-Sod2 Ϫ/Ϫ mice. Using an enzymatic histochemical analysis for SDH, we assessed biochemical activity of Complex II in cardiac muscle and skeletal muscle of 15-week-old H/M-Sod2 Ϫ/Ϫ mice and compared them with those of control mice. In the cardiac muscle of H/M-Sod2 Ϫ/Ϫ mice, the activity of Complex II was hardly detected except in endothelial cells of intramuscular vessel walls (Fig. 7A). Likewise, we detected no enzymatic activity of Complex II in the skeletal muscle of H/M-Sod2 Ϫ/Ϫ mice (data not shown), whereas a strong enzymatic staining of SDH (Complex II) activity was detected in the cardiac and skeletal muscles of control mice. We also assessed the activity of Complex IV, COX, in the cardiac muscle (Fig. 7A) and skeletal muscle (data not shown) of H/M-Sod2 Ϫ/Ϫ mice. The enzymatic activity of Complex IV in H/M-Sod2 Ϫ/Ϫ mice showed a strong staining that was comparable with control mice. The data clearly demonstrated a selective loss of enzymatic activity of Complex II but not of Complex IV in the cardiac and skeletal muscles of H/M-Sod2 Ϫ/Ϫ mice.
In order to biochemically confirm the suppression of OXPHOS, we isolated mitochondria from heart, muscle, and liver of H/M-Sod2 Ϫ/Ϫ mice as well as control mice. We then assessed NADH-cytochrome c reductase (Complex I ϩ III) activity and succinate-cytochrome c reductase (Complex II ϩ III) activity of H/M-Sod2 Ϫ/Ϫ mice. NADH-cytochrome c reductase activity in the heart and skeletal muscle was signifi-  ϭ 50 m). B and C, OXPHOS enzyme activities of heart, TA muscle, and liver mitochondria in H/M-Sod2 Ϫ/Ϫ (n ϭ 3) or control mice (n ϭ 3) at 15 weeks of age. Activities of OXPHOS Complex I ϩ III (B) and Complex II ϩ III (C ) are shown (*, p Ͻ 0.05; **, p Ͻ 0.005; ***, p Ͻ 0.0005). D, production of ATP in heart and liver mitochondria. ATP production in heart mitochondria from H/M-Sod2 Ϫ/Ϫ (n ϭ 5) was lower than that from control mice (n ϭ 4; *, p Ͻ 0.05). In liver mitochondria, ATP production was not significantly different between H/M-Sod2 Ϫ/Ϫ (n ϭ 3) and control mice (n ϭ 3). Results are represented by mean Ϯ S.E. E, Western blot analysis of proteins extracted from heart of 15-week-old H/M-Sod2 Ϫ/Ϫ using antibodies against OXPHOS subunits and actin. F, transcriptional alterations of Sdha, Sdhb, Ucp3, Abcc9, Ace, periostin, Itgb1bp3, and actin in the hearts (n ϭ 3).
cantly inhibited in H/M-Sod2 Ϫ/Ϫ mice compared with control mice (53.6 and 47.2% of control, respectively) (Fig. 7B). Furthermore, the succinate-cytochrome c reductase activity in heart and skeletal muscle of H/M-Sod2 Ϫ/Ϫ mice exhibited even more extensive reductions (29.5 and 4.7% of control, respectively) (Fig. 7C). We also measured the production of ATP in heart and liver mitochondria. Mitochondria isolated from the heart of H/M-Sod2 Ϫ/Ϫ mice generated less ATP (52.6 Ϯ 2.9% of that from control mice) (Fig. 7D). In the liver, we failed to detect any difference in ATP production between H/M-Sod2 Ϫ/Ϫ mice and control mice (Fig. 7D). As shown in Fig. 7E, immunoblot analyses on the expression of the SDHA and SDHB subunit from Complex II revealed significant reductions in the heart of H/M-Sod2 Ϫ/Ϫ mice, whereas we detected moderate suppression of Complex I ␣9, Rieske iron-sulfur protein (FeS) and Core I subunit of Complex III, and ␣ and ␤ subunits of Complex V and detected no suppression of COX I, a component of Complex IV (Fig. 7E). In order to clarify whether the suppression of the subunits is due to the transcriptional down-regulation or the posttranscriptional alterations of the subunits, we investigated the gene expression by RT-PCR. As shown in Fig. 7F, we detected comparable amounts of transcripts for sdha and sdhb in both H/M-Sod2 Ϫ/Ϫ and control mice. Thus, these results suggested that OXPHOS suppression was caused by selective posttranscriptional modifications of specific enzymes in the mitochondrial respiratory chain.
Identification of Up-and Down-regulated Genes in Mn-SODdeficient Heart-In order to investigate the transcriptional alterations in the hearts with dilatation of developed H/M-Sod2 Ϫ/Ϫ mice (15 weeks of age), we carried out an Affymetrix microarray analysis. We found 419 genes up-regulated and 118 genes down-regulated with transcriptional alterations of more than 2-fold compared with the control. The genes with more than 4-fold alternations are presented in biological categories ( Table 2). The transcriptional up-regulation of genes for cell growth, cell maintenance, cell adhesion, and organogenesis suggested active remodeling processes in fibrotic myocardial degeneration, whereas the transcriptional down-regulation of Ucp3 (uncoupling protein 3) and Fbp2 (fructose bisphosphatase 2). Interestingly, Itgb1bp3, a modulator of muscle proliferation and differentiation (30), was up-regulated 52-fold. However, transcripts for other members of Itgb1bp, Itgb1bp1 and Itgb1bp2 (melusin), were not changed. Melusin was associated with mechanical stress signaling in the heart (31). This suggests that Itgb1bp3 is a specific biomarker for cardiac oxidative stress ( Table 2). We failed to detect the transcriptional upor down-regulation of antioxidant enzymes, such as Cu/Zn-SOD, extracellular SOD, and catalase, suggesting that the Mn-SOD system was not compensated by another antioxidant system (data not shown). Next, we investigated the transcriptional alterations of candidate genes that had been reported in heart failure and/or dilated cardiomyopathy. Of more than 30 candidates, six genes were up-regulated, whereas only one gene, Abcc9, was significantly down-regulated 0.48-fold. The up-regulated transcripts included Ace, Acta1 (skeletal muscle actin ␣1), Gbe1 (glucan (1,4-␣-)branching enzyme 1), Lmna (lamin A), Nppa (natriuretic peptide precursor type A), and Nppb (natriuretic peptide precursor type B) in Mn-SOD-deficient heart (supplemental Table 1). Transcriptional alterations were also confirmed by RT-PCR (Fig. 7F).
Heart/Muscle Sod2 Ϫ/Ϫ Mice Showed Enhanced ROS Generation with Increased Lipid Peroxidation in Mitochondria-Because H/M-Sod2 Ϫ/Ϫ mice showed suppressed OXPHOS in the myocardium, we measured the formation of ROS, such as O 2 .
and H 2 O 2 . As shown in Fig. 8A, when we used 5 mM succinate as substrate, O 2 . formation in heart mitochondria from H/M-Sod2 Ϫ/Ϫ mice increased to 187% of control mice, and that in skeletal muscle mitochondria from H/M-Sod2 Ϫ/Ϫ mice increased to 493% of control mice. When we used 5 mM malate and 0.5 mM glutamate as substrates, O 2 . formation in heart mitochondria from H/M-Sod2 Ϫ/Ϫ mice increased to 161% of control mice (Fig. 8C). H 2 O 2 formation in skeletal muscle mitochondria from H/M-Sod2 Ϫ/Ϫ mice decreased to 52% of control mice with 5 mM succinate as a substrate (Fig. 8B). However, we failed to detect any difference in heart mitochondria between H/M-Sod2 Ϫ/Ϫ mice and control mice (Fig. 8, B and C). These results indicated that H/M-Sod2 Ϫ/Ϫ mice showed an enhanced In order to evaluate oxidative damage of H/M-Sod2 Ϫ/Ϫ mice, we measured the amount of malondialdehyde (MDA). We detected higher levels of MDA in heart mitochondria from H/M-Sod2 Ϫ/Ϫ mice than in those from control mice (Fig. 8D). In the cytoplasmic fraction of heart, however, we failed to detect any difference in the amount of MDA between H/M-Sod2 Ϫ/Ϫ mice and control mice. These results indicated that oxidative damages were specifically localized in mitochondria of H/M-Sod2 Ϫ/Ϫ mice.
Antioxidants Partially Ameliorated Symptoms of Heart/ Muscle-specific Mn-SOD-deficient Mice-In order to ameliorate the symptoms of H/M-Sod2 Ϫ/Ϫ mice, we intraperitoneally administered MnTBAP to mutant mice. We then assessed the heart weights and the cardiac function of 12-week-old H/M-Sod2 Ϫ/Ϫ mice. Although we failed to detect a significant pharmacological effect on the standardized heart weights of MnT-BAP-treated H/M-Sod2 Ϫ/Ϫ mice (Fig. 9A), we found a significant improvement in the cardiac function (Fig. 9B). The result suggested that MnTBAP reversed the weakened function of the failing heart, which was caused by the loss of Mn-SOD activity.
In the assessment of physical activities, we injected MnTBAP into 8-week-old H/M-Sod2 Ϫ/Ϫ mice for 5 weeks. We then compared the daily running distances of MnTBAP-treated mice with those of PBS-treated control mice. MnTBAP significantly improved the physical activity of H/M-Sod2 Ϫ/Ϫ mice, whereas PBS failed to improve the physical activity (Fig. 9C). Likewise, rotarod tasks with 14-week-old mutant mice showed a significant recovery on the administration of MnTBAP (Fig. 9D). These results suggested that the administration of MnTBAP significantly rescued the impaired cardiac contractility as well as the physical disabilities of the H/M-Sod2 Ϫ/Ϫ mice.

DISCUSSION
Total Mn-SOD KO mice have been independently reported by two laboratories (16,17). These authors showed that Mn-SOD-deficient mice die with complex pathologies, including growth immaturity, dilated cardiomyopathy, metabolic abnormalities, such as ketosis and lactic acidosis, steatosis or fatty change of the liver, and central nervous system damage within 3 weeks after birth (16,17). In the present study, we established mice carrying a specific deletion of the Mn-SOD gene in heart and muscle tissues to define the phenotypes of heart/musclespecific Mn-SOD deficiency from the complex pathologies observed in total Mn-SOD KO mice.
Mitochondrial O 2 . is formed from donation of an electron to molecular oxygen and then converted by Mn-SOD into H 2 O 2 .
Previous studies have suggested that H 2 O 2 plays a pathological . formation than that of control mice (heart, n ϭ 9; TA, n ϭ 8) with 5 mM succinate as substrate (**, p Ͻ 0.005). B, mitochondrial H 2 O 2 formation was measured using Amplex red and horseradish peroxidase. In skeletal muscle, H/M-Sod2 Ϫ/Ϫ (n ϭ 9) mice showed a higher level of H 2 O 2 formation than that of control mice at 8 weeks of age (n ϭ 9) with 5 mM succinate as substrate (*, p Ͻ 0.05; **, p Ͻ 0.0005). However, H 2 O 2 formation was not significantly different between H/M-Sod2 Ϫ/Ϫ (n ϭ 9) and control (n ϭ 9) mice in heart and liver. C, mitochondrial O 2 . formation in the hearts of H/M-Sod2 Ϫ/Ϫ and control mice at 15 weeks of age. Heart mitochondria from H/M-Sod2 Ϫ/Ϫ (n ϭ 9) mice showed a higher level of O 2 . formation than that of control mice (n ϭ 9) with 5 mM malate and 0.5 mM glutamate as substrates (*, p Ͻ 0.05). Mitochondria H 2 O 2 formation was measured using homovanilic acid as fluorophore in a peroxidase-catalyzed reaction. H 2 O 2 formations were not significantly different between H/M-Sod2 Ϫ/Ϫ (n ϭ 3) and control (n ϭ 3) mice. D, measurement of lipid peroxidation in mitochondria and cytoplasm fraction prepared from heart. Heart mitochondria from H/M-Sod2 Ϫ/Ϫ mice (n ϭ 3) showed a higher MDA level than those from control mice (n ϭ 3) (*, p Ͻ 0.05). All data are represented by the mean Ϯ S.E. role in heart failure (32,33 However, an increased level of ATP is a requisite to the process of apoptotic cell death (36). Leist et al. (37) have reported that depletion of cellular ATP could switch the type of cell death from apoptosis to necrotic cell death. Apoptosis is characterized by activation of caspase-3 and internucleosomal DNA fragmentations. Thus, the level of ATP is critical to the activation of caspase-3 (36). In the present study, dysfunction of mitochondrial respiratory chain led to reduction of ATP production (Fig. 7D) as well as ATP content (Fig. 6F), which could account for the absence of apoptosis in H/M-Sod2 Ϫ/Ϫ mice. Spontaneous apoptosis was reported to be observed in heterozygous Mn-SODdeficient mice (38). We speculate that the apoptosis occurred because there was sufficient ATP available to induce it.
In the heart mitochondria, H/M-Sod2 Ϫ/Ϫ mice showed significantly reduced ATP production (Fig. 7D). Two possibilities for the reduction of ATP production were suggested. Brand et al. reported that excess O 2 . generation in mitochondria uncoupled the respiration by activation of UCPs, which resulted in decreased ATP production (39). In microarray analysis, however, we failed to show up-regulation of Ucp1 and Ucp2 genes in the hearts of H/M-Sod2 Ϫ/Ϫ mice (data not shown). Inversely, the Ucp3 gene was markedly down-regulated 0.08-fold in hearts of H/M-Sod2 Ϫ/Ϫ mice compared with control mice ( Table 2). Another possibility is global reduction in mitochondrial respiration. In the present study, we showed marked reduction of Complex II (SDH) subunits and moderate suppression of subunits in Complexes I, III, and V (Fig. 7E). Therefore, it is suggested that reduced ATP production in mutant mitochondria is due to a decrease in mitochondrial respiration associated with suppression of OXPHOS proteins in H/M-Sod2 Ϫ/Ϫ mice. In the hearts of H/M-Sod2 Ϫ/Ϫ mice, Complex II (SDH) was selectively inactivated among mitochondrial respiratory enzymes. As presented under "Results," protein expression of SDHA and SDHB decreased significantly without down-regulation of mRNA (Fig. 7, E and F). The result suggested the pathological role of oxidative stress in the modification of the enzymes, which might induce the degradation of SDH protein.
It implies that Complex II is specifically vulnerable to O 2 . among mitochondrial respiratory chain complexes. It has been reported that oxidative stress activates redoxsensitive transcription factors, such as nuclear factor-B (NF-B) and activator protein-1 (40). NF-B is also known to be a key molecule of myocardial ischemia-reperfusion injury that generated elevated levels of free radicals (41). In this context, we investigated the expression of NF-B and its signaling molecules, such as inducible NO synthase, heme oxygenase-1, and interleukin-1␤. In microarray analysis, we could not detect any up-or down-regulation of these transcripts between H/M-Sod2 Ϫ/Ϫ and control mice (data not shown). Based on the transcriptional analysis, we speculated that NF-B is hardly attributable to the development and progression of dilated cardiomyopathy in H/M-Sod2 Ϫ/Ϫ mice.
Should free radicals play a pivotal role in the pathogenesis of dilated cardiomyopathy, dietary supplementation of anti-oxidants would confer a nutritional benefit. Broqvist et al. (42), however, reported that such clinical trials failed to show a significant benefit for dilated cardiomyopathy. In the present study, we investigated whether a diet high in glucose improved the physical activities of the mutant mice. The result showed that a high glucose diet did not improve physical activity as assessed using the running wheel apparatus (data not shown). Since some Mn-SOD mimetics, such as MnTBAP (43), EUK-8, and EUK-134 (44,45), showed antioxidant effects in model organisms, such as C. elegans and mice, we administered MnTBAP to the H/M-Sod2 Ϫ/Ϫ mice. The results revealed that MnTBAP ameliorates the phenotypical symptoms, physical disabilities, and cardiac contractility, suggesting that free radicals play an important role in the progression of dilated cardiomyopathy. Furthermore, antioxidant regimens can prevent the further progression of heart failure. Since MnTBAP only partially rescued the phenotypes, oxidative stress would have caused irreversible damage that could not be rescued using antioxidants. This suggests that patients with genetic susceptibility to heart failure might benefit from treatment with antioxidants in the early presymptomatic stage. In conclusion, we presented a murine model of heart/muscle-specific oxidative stress and devised a strategy for preventive medicine with a therapeutic treatment for age-related or oxidative stress-dependent heart diseases. Furthermore, we demonstrated the efficacy and limitations of antioxidant SOD mimetics. Future studies with our model mice should provide better ways to develop novel medicine or new dietary supplements for the improvement of physical disabilities as well as the quality of senescent life.