Heart Failure Affects Mitochondrial but Not Myofibrillar Intrinsic Properties of Skeletal Muscle
Background—Congestive heart failure (CHF) induces abnormalities in skeletal muscle that are thought to in part explain exercise intolerance. The aim of the present study was to determine whether these changes actually result in contractile or metabolic functional alterations and whether they are muscle type specific.
Methods and Results—With a rat model of CHF (induced by aortic banding), we studied mitochondrial function, mechanical properties, and creatine kinase (CK) compartmentation in situ in permeabilized fibers from soleus (SOL), an oxidative slow-twitch muscle, and white gastrocnemius (GAS), a glycolytic fast-twitch muscle. Animals were studied 7 months after surgery, and CHF was documented on the basis of anatomic data. Alterations in skeletal muscle phenotype were documented with an increased proportion of fast-type fiber and fast myosin heavy chain, decreased capillary-to-fiber ratio, and decreased citrate synthase activity. Despite a slow-to-fast phenotype transition in SOL, no change was observed in contractile capacity or calcium sensitivity. However, muscles from CHF rats exhibited a dramatic decrease in oxidative capacities (oxygen consumption per gram of fiber dry weight) of 35% for SOL and 45% for GAS (P<0.001). Moreover, the regulation of respiration with ADP and mitochondrial CK and adenylate kinase was impaired in CHF SOL. Mitochondrial CK activity and content (Western blots) were dramatically decreased in both muscles.
Conclusions—CHF results in alterations in both mitochondrial function and phosphotransfer systems but unchanged myofibrillar function in skeletal muscles, which suggests a myopathy of metabolic origin in CHF.
Muscular fatigue and reduced exercise tolerance are major symptoms of congestive heart failure (CHF). It is well accepted that central hemodynamics correlate poorly with exercise capacity in CHF. Attention has therefore been focused on peripheral factors such as alterations in skeletal muscle perfusion and intrinsic skeletal muscle abnormalities. Indeed, numerous studies in patients1 2 3 4 5 and animal models of heart failure6 7 8 9 10 11 have described muscle histological abnormalities, including fiber atrophy, fiber-type transformation toward more fast phenotype, and reduced oxidative enzyme activity. However, whether these abnormalities result in functional alterations of either the contractile machinery or sites of energy production has never been directly assessed.
31P NMR spectroscopy has revealed increased phosphocreatine (PCr) depletion and decreased intracellular pH during exercise in humans.12 PCr and creatine kinase (CK) are involved in the fine regulation between energy production and energy utilization in muscle cells. Generalized quantitative and functional alterations of the CK system are a hallmark of failing myocardium that contributes to alterations in intracellular energy fluxes and calcium homeostasis.13 14 Recently, a decreased content of mitochondrial CK (mi-CK) was observed in skeletal muscle of patients with CHF,15 although the CK function has never been assessed in skeletal muscle during CHF. Moreover, skeletal muscle in humans is composed of a mixture of fibers from fast glycolytic to slow oxidative, including intermediary fibers. Another important question, therefore, concerns the possible fiber or muscle type specificity of CHF-induced abnormalities.
The goal of the present study was therefore, with an animal model of prolonged chronic heart failure, to examine whether the morphological and biochemical changes observed in skeletal muscle will result in intrinsic contractile or metabolic alterations. Selective permeabilization of cellular membranes with detergents offers the unique opportunity to study in situ the myofibrillar and mitochondrial intrinsic properties, as well as CK compartmentation.16 We chose to examine the soleus (SOL), an oxidative slow-twitch muscle type, and the white gastrocnemius (GAS), a glycolytic fast-twitch muscle type, to also determine whether functional and biochemical changes might be fiber type specific.
Rat Model of CHF
Weaned male Wistar rats (60 to 70 g) were randomly assigned to 1 of 2 groups. Aortic stenosis was created by placing a stainless steel hemoclip of 0.6-mm internal diameter on the ascending aorta via a thoracic incision (CHF group) according to Feldman et al17 as previously described.14 Age-matched control animals underwent the same procedure without placement of the clip (sham group). This investigation was carried out in accordance with the Helsinki Recommendations for Humane Treatment of Animals During Experimentation.
At 7 months after surgery, 7 CHF and 7 sham animals were anesthetized with an intraperitoneal injection of urethane (0.2 g/100 g), and the right and left SOL and superficial portions of GAS were isolated. A portion of the muscles were rapidly frozen for biochemical determinations.
Functional Properties of Mitochondria and Bound CK and Adenylate Kinase
Respiratory parameters of the total mitochondrial population were studied in situ in fresh saponin-skinned fibers as previously described14 and determined with a Clark electrode (Strathkelvin Instruments) in an oxygraphic cell containing 3 mL respiration solution (see later) at 22°C with continuous stirring. Respiration rates were expressed as μmol O2 · min−1 · g dry wt−1. Respiration solution was calculated with the computer program of Fabiato18 and contained (in mmol/L) EGTA-CaEGTA buffer 10 (free Ca2+ concentration 100 nmol/L), MgCl2 1, taurine 20, dithiothreitol 0.5, imidazole 20 (pH 7.1), ionic strength 160 (potassium methane sulfonate), glutamate 5, malate 2, and phosphate 3 and 2 mg/mL fatty acid–free BSA. The ADP-stimulated respiration (VADP) above basal oxygen consumption (V0) was plotted as a function of [ADP] with or without CK (20 mmol/L). The apparent Km values for ADP and VADP were calculated with nonlinear fit of the Michaelis-Menten equation. The maximal respiration rate (Vmax) was (VADP+V0). The acceptor control ratio (ACR) was Vmax/V0. The functional activity of adenylate kinase (AK) was evaluated in the presence of 0.2 mmol/L ATP on the basis of the percent increase in the respiration rate after the addition of 2 mmol/L AMP (VAMP%). From 1 to 3 determinations were made for each animal.
Muscle fiber bundles were dissected from SOL and GAS muscles in a zero-Ca2+ Krebs’ solution (pH 7.4) and permeabilized with 1% Triton X-100 in a relaxing solution. Bundles were mounted between a vibrator and a force transducer (model AE 801; SensoNor Microelectroniks) as previously described.19 Sarcomere length was measured with laser diffraction and adjusted to 2.5 to 2.6 μm. Solutions were calculated with the computer program of Fabiato18 and contained (in mmol/L) EGTA 10 (pCa 9 to 4.5), imidazole 30 (pH 7.1), Na+ 30.6, Mg2+ 3.16, dithiothreitol 0.3, MgATP 3.16, PCr 12, and ionic strength 160 (potassium acetate). Data for each bundle were fit with nonlinear fit of the Hill equation: T=Ln/(K+Ln), where L is the calcium concentration, T is the relative tension, K is a constant, and n is the Hill coefficient.
Frozen tissue samples were weighed and homogenized in ice-cold buffer (50 mg/mL) containing (in mmol/L) HEPES 5 (pH 8.7), EGTA 1, dithiothreitol 1, and MgCl2 5 and 0.1% Triton X-100 and incubated for 60 minutes at 0°C to ensure complete enzyme extraction. Enzyme activities were determined at 30°C, pH 7.5, as previously described with coupled enzyme systems.14 CK and LDH isoenzymes were separated with agarose (1%) gel electrophoresis performed at 200 V for 90 minutes and revealed through incubation of the gels with a coupled enzyme system (CK) or commercial revelation system (LDH reagent kit; Sigma Chemical Co).
Native myosin was extracted as previously described.19 Myosin heavy chain (MHC) isoforms were separated with PAGE (90 V, 22 hours, 3°C). Immunohistochemical studies were carried out on the midbelly portion of SOL and GAS. Serial transverse sections (10 μm thick) were cut out at −20°C on a cryostat and incubated for 1 hour at 37°C with specific antibodies raised against (1) slow-type I (Novocastra, reference NCL-MHCS), (2) all adult fast and developmentally regulated epitopes but not slow myosin (MY-32; Sigma Chemical Co), (3) fast-type IIa (SC-71), (4) slow- and fast-type IIa and type IIb but not type IIx MHC (BF-35), (5) or fast-type IIb MHC isoforms (BF-F3). The avidin-biotin immunohistochemical procedure was used for localization of the antigen-antibody binding (Vector Laboratories). A sample of ≈400 fibers was randomly selected and classified according to the staining profile with a microscope linked to a computer-based image analysis system (Visiolab 200; Nikon-France). Negative control slides with omission of the primary antibodies were randomly included in the immunostaining procedures.
Western Blot Analysis
Mouse mi-CK antibody (kind gift of Drs Z. Khuchua and W. Qin, Washington University, St Louis, Mo) was produced in rabbit against mouse whole recombinant sarcomeric mi-CK.20 Specificity of the antibody and Western blot analyses were performed as previously described.14 Briefly, 1.25 SOL and 5 GAS μg protein extracts were separated, subsequently transferred to Hybond nitrocellulose membranes (Amersham), and incubated with mi-CK antibody for 2 hours. Membranes were revealed with ECL+ chemiluminescent substrate (Amersham). Light emission was detected with autoradiography, quantified with an image analysis system (Bio-Rad), and normalized to SOL extract from sham group.
Capillaries were visualized with an acidic ATPase reaction after preincubation at pH 421 and identified with a microscope linked to a computer-based image analysis system (Visiolab 200; Nikon-France). From 4 to 6 areas were selected on each sample (total area A). The capillary bed was appraised by (1) the capillary density, calculated as the number of capillaries in A divided by the area of A, and (2) the capillary-to-fiber ratio, determined as capillary density normalized by the mean number of fibers/mm2.
All data are expressed as mean±SEM. Data were statistically evaluated with a Student’s t test between sham and CHF. A 2-way ANOVA, followed by Newman-Keuls post hoc test when appropriate, was used to assess the main effects of muscle type (SOL versus GAS) and condition (sham versus CHF). Values of P<0.05 were considered significant.
CHF animals showed decreased body weight but normal skeletal growth on the basis of a similar tibia length (Table 1⇓). Increased absolute and relative weights of the heart, both left and right ventricles, and the lung in the CHF group indicated severe heart failure. Anatomic evidence of cardiac decompensation (ascites, congestion, pleural effusion, and edema), as well as our previous hemodynamic data (eg, increased left ventricular end-diastolic pressure from 10±4 to 62±18 mm Hg; see De Sousa et al14 for details), confirmed the occurrence of severe CHF in this model.
Morphological and Mechanical Studies
CHF induced a significant absolute and relative decrease in SOL weight (Table 2⇓). The proportion of fibers that expressed the slow MHC in CHF SOL decreased by 10%, whereas that of fibers that expressed fast MHC exhibited a 6-fold increase, only due to an increase in IIa MHC expression. Fast IIb MHC was not detectable in CHF SOL. The proportion of slow MHC isoforms decreased and fast MHC-IIa increased in CHF SOL (Table 2⇓). The appearance of IIx MHC in CHF SOL was not significant. GAS exhibited exclusively fast IIx and IIb MHC isoforms. No significant changes were observed in CHF.
The number of capillaries per fiber and capillary density were higher in SOL than GAS muscle. In both CHF skeletal muscles, the number of capillaries per fiber was decreased, but capillary density remained unchanged.
Myofibrillar function was studied in Triton X-100–treated fiber bundles to assess whether changes in contractile pattern will result in alteration in the whole muscle function. Active tension was plotted as a function of pCa in SOL (Figure 1A⇓) and GAS (Figure 1B⇓) from control and CHF rats. Resting and active tensions were significantly higher in GAS than in SOL, as well as slope coefficient (nH), whereas calcium sensitivity (pCa50) was lower (Table 3⇓). None of these mechanical parameters were altered in the CHF group.
The activity of AK, a phosphotransfer enzyme, higher in GAS than SOL, was unchanged in both CHF muscles (Table 4⇓). Compared with SOL, GAS exhibited a clear glycolytic profile as shown by higher total LDH, PK, and phosphoglycerate kinase (PGK) activities; lower activity of CS; and absence of the aerobic subunit (H subunit) of LDH. Activities of PK and PGK, the 2 steps of glycolytically produced ATP, were maintained in both CHF muscles. Total LDH activity was not changed in CHF muscles, but the ratio of H/M isoforms decreased in CHF SOL. CS activity was depressed in both muscles.
Both skeletal muscles expressed MM-CK and mi-CK isoenzymes only (Table 5⇓). GAS showed a typical fast glycolytic profile as indicated by higher total and MM-CK activities and lower content and activity of mi-CK. Total CK and MM-CK activities were unchanged in CHF GAS but decreased in CHF SOL. The activity of mi-CK decreased dramatically by 45% in SOL and 83% in GAS. The mi-CK–to–citrate synthase ratio was similarly decreased. Western blot analysis showed that this decrease could be accounted for by a 34% and 45% decrease in mi-CK protein content in CHF SOL and GAS, respectively.
Mitochondrial Function and CK
Oxygen consumption of permeabilized preparations from sham and CHF rats was recorded as a function of ADP concentration with or without creatine (Figure 2⇓). Respiration rates were lower in GAS than in SOL, whereas the ACRs were comparable (Table 6⇓). In CHF rats, oxidative capacities decreased by ≈35% and ≈45% in SOL and GAS, respectively. Because the ACRs were high and preserved in CHF, this decrease suggested a decreased amount of mitochondria without changes in the oxidation-phosphorylation coupling. SOL was characterized by a low sensitivity of respiration for extramitochondrial ADP and a decreased apparent Km value for ADP in the presence of creatine, whereas for GAS, the apparent Km value for ADP was very low and unchanged by the addition of creatine. In the absence of creatine, the Km value for ADP was significantly lower in CHF SOL and unchanged in CHF GAS. The addition of creatine decreased the Km value for ADP to a higher extent in CHF SOL than in control. The stimulation of respiration by AMP in the presence of ATP was 40% in control and only 22% in CHF SOL (P<0.01), showing an impairment of mitochondrial AK efficacy.
The present study shows that the changes in contractile phenotype of skeletal muscles observed in CHF do not result in alterations in intrinsic mechanical properties of myofilaments. On the contrary, mitochondrial respiration and energy transfer systems (CK) are quantitatively and qualitatively altered in CHF muscles. These alterations are more relevant to a metabolic myopathy in heart failure than a simple phenoconversion. The fact that mitochondria of both fast glycolytic and slow oxidative muscles are affected suggests that a central mechanism might be involved.
This experimental model of heart failure has been well characterized14 17 and reproduced the main skeletal muscle abnormalities previously described in animal studies or in patients with CHF.2 3 5 7 9 11 22 We report here a reduced number of capillaries and changes in MHC expression in SOL from CHF rats that consist of decreased type I MHC expression and increased fast-type IIa MHC expression. Fibers generally express a whole set of contractile proteins of a specific type, including regulatory proteins of the thin filament. This results in different intrinsic mechanical properties such as a higher calcium sensitivity and lower Hill coefficient in slow versus fast fibers (for a review, see Schiaffino and Reggiani23 ). Despite changes in myosin expression and decreased expression of skeletal actin,10 skinned fibers of SOL muscle of CHF rats exhibited mechanical properties identical to those of control fibers. Although nothing is known concerning possible changes in troponin isoform expression in skeletal muscle in CHF, these changes are not important enough to modify contractile capacity and sensitivity to calcium.
Metabolism and CK System
Skeletal muscle phenotype is also characterized by the type of energy metabolism adapted to function. Fast fibers have a predominantly glycolytic metabolism, whereas slow fibers are mainly oxidative. In either muscle type, CHF did not result in altered glycolytic enzyme activity, which is in agreement with studies in animals or patients.1 2 6 11 It should be mentioned that the slight decrease in LDH isoenzyme ratio nevertheless suggests an increase in anaerobic metabolism in CHF SOL. Energy transfer systems are important in the adjustment of energy production to demand. CK fulfills specific roles in glycolytic and oxidative muscles.24 25 Fast glycolytic muscles have a high activity of both total and the cytosolic isoform of CK (MM-CK) as correlated to the speed of contraction,26 whereas slow oxidative muscles exhibit a lower total activity of CK and a higher amount of mi-CK, which provides an efficient system of energy transfer from mitochondria to ATPases. This study shows that the CK system is clearly altered in skeletal muscle from rats with severe CHF. However, as a result of a slow-to-fast shift in fiber type in SOL, an increase in MM-CK was expected rather than a decrease.19 This decrease in MM-CK is reminiscent of what is observed in the failing myocardium of humans and animals13 14 and seems to be specific to heart failure. In addition, a drop in mi-CK that exceeds the decrease in mitochondrial mass was observed in both muscle types, extending the observations made by Hambrecht et al15 in patients. Mi-CK decrease could thus be proposed as a generalized marker of the severity of metabolic alterations in myocardium and skeletal muscles of patients with CHF (see Nascimben et al13 and De Sousa et al14 and references therein).
Early PCr depletion, decreased intracellular pH, increased lactate production at exercise, and alterations in mitochondrial volume density and enzyme activities1 2 3 4 6 8 9 10 11 12 have supported the proposal that oxidative skeletal muscle metabolism is impaired in CHF, although decreased mitochondrial enzyme activities are not always reported and could be enzyme or muscle dependent.1 11 22 These results provide the first functional evidence for a marked fall in oxidative capacity, affecting both oxidative and glycolytic muscles.
Moreover, we could observe qualitative changes in mitochondrial respiration specifically in oxidative muscle. CHF resulted in a decrease in the control of respiration by mitochondrial kinases (AK and CK). Mitochondrial respiration in oxidative muscles is preferentially regulated by intramitochondrially produced ADP through phosphokinases, whereas such fine control is absent in glycolytic muscle.16 24 This regulation provides a specialized system through which a cytosolic metabolic pathway is linked to mitochondrial respiration.27 Whether this increase in ADP sensitivity could be a compensatory mechanism for the drop in mi-CK activity and content to preserve mitochondrial function is at present only speculative.
The general conclusion of the present study is that heart failure markedly affected the mitochondrial capacity and regulation rather than intrinsic contractile machinery of skeletal muscles. Many studies, including the close relation between peak V̇o2 and mitochondrial enzyme activity or volume,3 4 28 have pointed out that impaired aerobic metabolism may play a role in the limitation of systemic exercise capacity. The decreased oxidative capacity and altered mitochondrial regulation and energy transfer demonstrated here could be the mechanistic basis for decreased oxygen utilization and exercise capacity in CHF. Because skeletal muscle abnormalities have some similarities with deconditioning, reduced physical activity of patients with CHF has been proposed as a possible cause of these alterations.1 3 However, several observations have raised the possibility of a specific myopathy in this pathology. In some cases, a decrease in (or even disappearance of) type I MHC expression may far exceed changes attributable to exercise deconditioning.5 Considering animal models of heart failure, reduced activity cannot explain skeletal muscle abnormalities.9 Moreover, in a model of severe muscle deconditioning induced by hindlimb suspension, oxidative metabolism and mitochondrial properties were preserved in contrast to what is observed in heart failure.19 This argues the case for a specific metabolic myopathy in heart failure.
The fact that decreased oxidative capacity and altered mitochondrial regulation can be encountered in cardiac slow-oxidative and fast-glycolytic muscles suggests that central mechanism or systemic factors are involved. Alterations in the neurohumoral system or circulating factors such as cytokines and tissue necrosis factor can affect skeletal muscle. Recently, a negative correlation was found among the expression of mi-CK in skeletal muscle, exercise capacity, and the expression of inducible NO synthase, suggesting that inducible NO synthase may be involved in exercise intolerance of patients with CHF.15
Dr Ventura-Clapier is supported by Centre National de la Recherche Scientifique. This work was supported by a PROGRES grant from INSERM and a grant from Association Française contre les Myopathies.
- Received February 11, 2000.
- Revision received April 26, 2000.
- Accepted May 9, 2000.
- Copyright © 2000 by American Heart Association
Mancini DM, Coyle E, Coggan A, et al. Contribution of intrinsic skeletal muscle changes to 31P NMR skeletal muscle metabolic abnormalities in patients with chronic heart failure. Circulation. 1989;80:1338–1346.
Sullivan MJ, Green HJ, Cobb FR. Skeletal muscle biochemistry and histology in ambulatory patients with long-term heart failure. Circulation. 1990;81:518–527.
Drexler H, Riede U, Munzel T, et al. Alterations of skeletal muscle in chronic heart failure. Circulation. 1992;85:1751–1759.
Arnolda L, Brosnan J, Rajagopalan B, et al. Skeletal muscle metabolism in heart failure in rats. Am J Physiol. 1991;261:H434–H442.
Sabbah HN, Hansen-Smith F, Sharov VG, et al. Decreased proportion of type I myofibers in skeletal muscle of dogs with chronic heart failure. Circulation. 1993;87:1729–1737.
Simonini A, Long CS, Dudley GA, et al. Heart failure in rats causes changes in skeletal muscle morphology and gene expression that are not explained by reduced activity. Circ Res. 1996;79:128–136.
Delp MD, Duan C, Mattson JP, et al. Changes in skeletal muscle biochemistry and histology relative to fiber type in rats with heart failure. J Appl Physiol. 1997;83:1291–1299.
Nascimben L, Ingwall JS, Pauletto P, et al. Creatine kinase system in failing and nonfailing human myocardium. Circulation. 1996;94:1894–1901.
De Sousa E, Veksler V, Minajeva A, et al. Subcellular creatine kinase alterations: implications in heart failure. Circ Res. 1999;85:68–76.
Feldman AM, Weinberg EO, Ray PE, et al. Selective changes in cardiac gene expression during compensated hypertrophy and the transition to cardiac decompensation in rats with chronic aortic banding. Circ Res. 1993;73:184–192.
Schiaffino S, Reggiani C. Molecular diversity of myofibrillar proteins: gene regulation and functional significance. Physiol Rev. 1996;76:371–422.
Veksler VI, Kuznetsov AV, Anflous K, et al. Muscle creatine kinase-deficient mice, 2: cardiac and skeletal muscles exhibit tissue-specific adaptation of the mitochondrial function. J Biol Chem. 1995;270:19921–19929.
Brdiczka D, Wallimann T. The importance of the outer mitochondrial compartment in regulation of energy metabolism. Mol Cell Biochem. 1994;133:69–83.