Mitochondria as a source of mechanical signals in cardiomyocytes.

AIMS: The myofibrillar and nuclear compartments in cardiomyocytes are known to be sensitive to extracellular mechanical stimuli. Recently, we have shown that alterations in the mitochondrial ionic balance in cells in situ are associated with considerably increased mitochondrial volume. Theoretically, this swelling of mitochondria could impose mechanical constraints on the myofibrils and nuclei in their vicinity. Thus, we studied whether modulation of mitochondrial volume in cardiomyocytes in situ has a mechanical effect on the myofibrillar and nuclear compartments. METHODS AND RESULTS: We used the measurement of passive force developed by saponin-permeabilized mouse ventricular fibres as a sensor for compression of the myofibrils. Osmotic compression induced by dextran caused an increase in passive force. Similarly, mitochondrial swelling induced by drugs that alter ionic homeostasis (alamethicin and propranolol) markedly augmented passive force (confirmed by confocal microscopy). Diazoxide, a mitochondrial ATP-sensitive potassium channel opener known to cause moderate mitochondrial swelling, also increased passive force (by 28 +/- 5% at 10% stretch, P < 0.01). This effect was completely blocked by 5-hydroxydecanoate (5-HD), a putative specific inhibitor of these channels. Mitochondrial swelling induced by alamethicin and propranolol led to significant nuclear deformation, which was visualized by confocal microscopy. Furthermore, diazoxide decreased nuclear volume, calculated using three-dimensional reconstructed images, in a 5-HD-dependent manner by 12 +/- 2% (P < 0.05). This corresponds to an increase in intracellular pressure of 2.1 +/- 0.3 kPa. CONCLUSION: This study is the first to demonstrate that mitochondria are able to generate internal pressure, which can mechanically affect the morphological and functional properties of intracellular organelles.

Local energetic regulation of sarcoplasmic and myosin ATPase is differently impaired in rats with heart failure.

Local control of ATP/ADP ratio is essential for efficient functioning of cellular ATPases. Since creatine kinase (CK) activity and mitochondrial content are reduced in heart failure (HF), and cardiomyocyte ultrastructure is altered, we hypothesized that these changes may affect the local energetic control of two major cardiac ATPases, the sarcoplasmic reticulum (SR) Ca2+-ATPase (SERCA) and the myosin ATPase. Heart failure was induced by aortic stenosis in rats. Electron microscopy confirmed that failing cardiomyocytes had intracellular disorganization, with fewer contacts between mitochondria and myofibrils. Despite normal SERCA protein content, spontaneous Ca2+ release measurements using Fluo-4 on saponin-permeabilized cardiomyocytes showed a lower SR loading in HF even when endogenous CK and mitochondria were fully activated. Similarly, in permeabilized fibres, SR Ca2+ loading supported by SR-bound CK and mitochondria was significantly reduced in HF (by 49% and 40%, respectively, 43% when both systems were activated, P < 0.05). Alkaline phosphatase treatment had no effect, but glycolytic substrates normalized calcium loading in HF to the sham level. The control by CK and mitochondria of the local ATP/ADP ratio close to the myosin ATPase (estimated by rigor tension) was also significantly impaired in HF fibres (by 32% and 46%, respectively). However, while the contributions of mitochondria and CK to local ATP regeneration were equally depressed in HF for the control of SERCA, mitochondrial contribution was more severely impaired than CK (P < 0.05) with respect to myofilament regulation. These data show that local energetic regulation of essential ATPases is severely impaired in heart failure, and undergoes a complex remodelling as a result of a decreased activity of the ATP-generating systems and cytoarchitecture disorganization.

AMP-activated protein kinase alpha2 deficiency affects cardiac cardiolipin homeostasis and mitochondrial function.

AMP-activated protein kinase (AMPK) plays an important role in controlling energy homeostasis and is envisioned as a promising target to treat metabolic disorders. In the heart, AMPK is involved in short-term regulation and in transcriptional control of proteins involved in energy metabolism. Here, we investigated whether deletion of AMPKalpha2, the main cardiac catalytic isoform, alters mitochondrial function and biogenesis. Body weight, heart weight, and AMPKalpha1 expression were similar in control littermate and AMPKalpha2(-/-) mice. Despite normal oxygen consumption in perfused hearts, maximal oxidative capacity, measured using saponin permeabilized cardiac fibers, was approximately 30% lower in AMPKalpha2(-/-) mice with octanoate, pyruvate, or glutamate plus malate but not with succinate as substrates, showing an impairment at complex I of the respiratory chain. This effect was associated with a 25% decrease in mitochondrial cardiolipin content, the main mitochondrial membrane phospholipid that is crucial for complex I activity, and with a 13% decrease in mitochondrial content of linoleic acid, the main fatty acid of cardiolipins. The decrease in cardiolipin content could be explained by mRNA downregulation of rate-limiting enzymes of both cardiolipin synthesis (CTP:PA cytidylyltransferase) and remodeling (acyl-CoA:lysocardiolipin acyltransferase 1). These data reveal a new role for AMPKalpha2 subunit in the regulation of cardiac muscle oxidative capacity via cardiolipin homeostasis.

Altered energy transfer from mitochondria to sarcoplasmic reticulum after cytoarchitectural perturbations in mice hearts.

Sarcoplasmic reticulum (SR) calcium pump function requires a high local ATP/ADP ratio, which can be maintained by direct nucleotide channelling from mitochondria, and by SR-bound creatine kinase (CK)-catalysed phosphate-transfer from phosphocreatine. We hypothesized that SR calcium uptake supported by mitochondrial direct nucleotide channelling, but not bound CK, depends on the juxtaposition of these organelles. To test this, we studied a well-described model of cytoarchitectural disorganization, the muscle LIM protein (MLP)-null mouse heart. Subcellular organization was characterized using electron microscopy, and mitochondrial, SR and myofibrillar function were assessed in saponin-permeabilized fibres by measuring respiration rates and caffeine-induced tension transients. MLP-null hearts had fewer, less-tightly packed intermyofibrillar mitochondria, and more subsarcolemmal mitochondria. The apparent mitochondrial Km for ADP was significantly lower in the MLP-null heart than in control (175 +/- 15 and 270 +/- 33 microM, respectively), indicating greater ADP accessibility, although maximal respiration rate, mitochondrial content and total CK activity were unaltered. Active tension in the myofibres of MLP-null mice was 54% lower than in controls (39 +/- 3 and 18 +/- 1 mN mm(-2), respectively), consistent with cytoarchitectural disorganization. SR calcium loading in the myofibres of MLP-null mice was similar to that in control myofibres when energy support was provided via Bound CK, but approximately 36% lower than controls when energy support was provided by mitochondrial (P < 0.05). Mitochondrial support for SR calcium uptake was also specifically decreased in the desmin-null heart, which is another model of cytoarchitectural perturbation. Thus, despite normal oxidative capacity, direct nucleotide channelling to the SR was impaired in MLP deficiency, concomitant with looser mitochondrial packing and increased nucleotide accessibility to this organelle. Changes in cytoarchitecture may therefore impair subcellular energy transfer and contribute to energetic and contractile dysfunction.

MLP accumulation and remodelling in the infarcted rat heart.

Mutation of cytoskeletal protein genes results in abnormal protein function and causes cardiomyopathy. We hypothesised that cardiac levels of cytoskeletal proteins, such as dystrophin, desmin and muscle LIM protein (MLP), would be altered during remodelling caused by myocardial infarction (MI). We measured left-ventricular morphology, function and cytoskeletal protein levels 10 weeks after coronary artery ligation or sham operation in male Wistar rats. Two-dimensional echocardiography revealed significant impairment of systolic function and decreased ejection fraction in infarcted hearts compared with sham (47+/-5% versus 73+/-4%), commensurate with the development of heart failure. Western blotting was used to measure levels of beta-myosin heavy chain (beta-MyHC), a marker of hypertrophy, and levels of dystrophin, desmin, MLP, beta-tubulin, utrophin and syncoilin, using GAPDH for normalization. Relative to shams, beta-MyHC and MLP levels were increased 1.9-fold and 1.7-fold, respectively, in infarcted rat hearts, whereas the levels of other cytoskeletal proteins were unchanged. Both MLP and desmin protein levels correlated negatively with ejection fraction, with the strongest relation between MLP and ejection fraction (r=-0.95, n=13, p<0.0001). This work suggests that MLP may play an important compensatory role in cardiac remodelling following MI.

Dystrophin- and MLP-deficient mouse hearts: marked differences in morphology and function, but similar accumulation of cytoskeletal proteins.

In humans, cytoskeletal dystrophin and muscle LIM protein (MLP) gene mutations can cause dilated cardiomyopathy, yet these mutations may have different effects in mice, owing to increased accumulation of other, compensatory cytoskeletal proteins. Consequently, we characterized left-ventricular (LV) morphology and function in vivo using high-resolution cine-magnetic resonance imaging (MRI) in 2- to 3-month old dystrophin-deficient (mdx) and MLP-null mice, and their respective controls. LV passive stiffness was assessed in isolated, perfused hearts, and cytoskeletal protein levels were determined using Western blot analyses. In mdx mouse hearts, LV-to-body weight ratio, cavity volume, ejection fraction, stroke volume, and cardiac output were normal. However, MLP-null mouse hearts had 1.2-fold higher LV-to-body weight ratios (P<0.01), 1.5-fold higher end-diastolic volumes (P<0.01), and decreased ejection fraction compared with controls (25% vs. 66%, respectively, P<0.01), indicating dilated cardiomyopathy and heart failure. In both models, isolated, perfused heart end-diastolic pressure-volume relationships and passive left-ventricular stiffness were normal. Hearts from both models accumulated desmin and beta-tubulin, mdx mouse hearts accumulated utrophin and MLP, and MLP-null mouse hearts accumulated dystrophin and syncoilin. Although the increase in MLP and utrophin in the mdx mouse heart was able to compensate for the loss of dystrophin, accumulation of desmin, syncoilin and dystrophin were unable to compensate for the loss of MLP, resulting in heart failure.