Mitochondria are essential for energy supply and cell signalling and may be triggers and effectors of cell death. Mitochondrial respiration is tightly controlled by the matrix Ca2+ concentration, which is beat-to-beat regulated by uptake and release mainly through the mitochondrial Ca2+ uniporter and Na+/Ca2+ exchanger, respectively. Recent studies demonstrate that mitochondrial Ca2+ uptake is more dependent on anatomo-functional microdomains established with the sarcoplasmic reticulum (SR) than on cytosolic Ca2+. This privileged communication between SR and mitochondria is not restricted to Ca2+ but may involve ATP and reactive oxygen species, which has important implications in cardiac pathophysiology. The disruption of the SR–mitochondria interaction caused by cell remodelling has been implicated in the deterioration of excitation–contraction coupling of the failing heart. The SR–mitochondria interplay has been suggested to be involved in the depressed Ca2+ transients and mitochondrial dysfunction observed in diabetic hearts as well as in the genesis of certain arrhythmias, and it may play an important role in myocardial reperfusion injury. During reperfusion, re-energization in the presence of cytosolic Ca2+ overload results in SR-driven Ca2+ oscillations that may promote mitochondrial permeability transition (MPT). The relationship between MPT and Ca2+ oscillations is bidirectional, as recent data show that the induction of MPT in Ca2+-overloaded cardiomyocytes may result in mitochondrial Ca2+ release that aggravates Ca2+ handling and favours hypercontracture. A more complete characterization of the structural arrangements responsible for SR–mitochondria interplay will allow better understanding of cardiac (patho)physiology but also, and no less important, should serve as a basis for the development of new treatments for cardiac diseases.
Mitochondria have long been recognized as critical players in eukaryotic cell metabolism and energetics, but it was not until recent years that they have re-emerged as determinants of cell death or survival, with great relevance in cardiovascular (patho)physiology. The possibility that they form local privileged communication with other cell organelles, in particular with endo(sarco)plasmic reticulum, has aroused a new area of interest among cardiovascular scientists.
Mitochondria made complexity possible
Energy suppliers and cell signalling mediators
The ability of cells to produce energy from atmospheric oxygen is possible only because of mitochondria, and this feature is in the basis of multicellular life. Electrons from dietary substrates are transferred by the respiratory complexes to the oxygen, generating an H+ electrochemical gradient (ΔΨm), whose energy is used by the F1Fo-ATP synthase to drive ATP synthesis. Remarkably, as the ΔΨm dissipates, i.e. under hypoxic conditions or when H+ influx follows an alternative route not coupled to energy production, mitochondria can act as ATP consumers by reversing the activity of ATP synthase into ATP hydrolase in an attempt to maintain their own ΔΨm.1 The endosymbiotic evolutionary origin of mitochondria may provide the clues to understanding their molecular response under this and other stressful conditions. Mitochondrial energy production requires the coordination of several sequential steps tightly interconnected and suited to cellular requirements. However, even under well-coupled respiration, up to 1% of electron flux results in an incomplete reduction of molecular oxygen and production of superoxide that is subsequently converted to hydrogen peroxide.2 Thus, mitochondria are the main source of reactive oxygen species (ROS) production within cells. Whereas a mild increase in oxidative stress acts as a cell signalling mechanism required to trigger several stress responses,3 the insidious increase in free radical production or a burst of oxidative damage may risk mitochondrial integrity and exacerbate cell damage during different processes, from ageing to ischaemia–reperfusion.2,4 Superoxide production is very sensitive to the ΔΨm, being strongly decreased by mild uncoupling. It has been proposed that the modulation of the inner membrane H+ permeability (by uncoupling proteins, mitochondrial ion channels, and fatty acids, among others) may be an important mechanism of cardioprotection due to its effect on ROS production by mitochondria.5
Contribution of mitochondria to ionic homeostasis
From the pioneering studies of Carafoli and Lehninger6 and Carafoli et al.,7 mitochondria have been recognized as organelles capable of accumulating large quantities of calcium. The physiological significance and the specific routes of mitochondrial calcium uptake pathways, however, have been much debated. Mitochondria take up calcium primarily through the calcium uniporter,8 whose isolation and purification has not yet been accomplished, in a manner that is dependent on ΔΨm at the physiological range of cytosolic calcium concentration. Calcium extrusion from the matrix may take place even after ΔΨm dissipation through the mitochondrial Na+/Ca2+ exchanger (mNCX),9 through reversal of the uniporter,10 or by permeability transition of mitochondrial membranes.11–13 Mitochondrial calcium controls the rate of energy production14 and plays an instrumental role in cell apoptosis and necrosis.15–17 Moreover, mitochondrial calcium uptake and release participates in the regulation of the amplitude and spatio-temporal pattern of intracellular calcium signals in different cell types, including contracting cardiomyocytes.8–20
Among the several mitochondrial membrane ion channels and exchangers, K channels (either ATP-sensitive or calcium-activated) have been extensively investigated for their potential cardioprotective roles. The opening of mitochondrial K-ATP channels21,22 or calcium-activated K channels23 has been shown to be beneficial for cell survival under certain stress conditions, including ischaemia–reperfusion and heart failure. The precise mechanism of this protective effect has not been elucidated, but it is postulated that the opening of these channels produces both an increase in mitochondrial volume24 and a certain degree of mitochondrial uncoupling which, in turn, reduces calcium overload.25 The regulation of matrix volume is crucial for the correct functioning of mitochondria and is accomplished indirectly by a number of specific exchangers, uniporters, and ion channels present in both membranes. Moderate matrix swelling has been shown to exert a protective function against myocardial ischaemic–reperfusion damage.24 In particular, voltage-dependent anion channel or mitochondrial porin, the most abundant protein of the outer membrane, is a large, water-filled pore that allows charged molecules to cross freely to the intermembrane space.26 In the last years, aquaporins have been found associated with mitochondrial inner membranes, where they may constitute the molecular pathway underlying osmotic movement of water to the matrix.27 Connexin 43, a protein constitutively forming sarcolemmal hemichannels and gap junction channels in ventricular myocytes, is also present at the inner mitochondrial membranes28,29 where it contributes to mitochondrial K+ influx.30 Resting intramitochondrial pH has been estimated at 7.7–8, is controlled by several inner membrane transporters and exchangers, and may influence cytosolic pH. Thus, reverse operation of FoF1 ATPase has been implicated in cytosolic acidification (concomitant with mitochondrial alkalinization), and represents a triggering event in the mitochondria-dependent pathway for caspase activation and apoptosis.31
Mitochondrial death pathways
Mitochondria are the central regulators of cell fate during exposure to stress, as they contain the specific and latent mechanisms for activating both the intrinsic apoptotic pathway and necrotic death in a variety of cell types.19,32 The intrinsic apoptotic pathway involves the activation and subsequent translocation to mitochondria of the members of the Bcl2 family (Bax, Bak, and Bid), following the canonical mechanism of programmed cell death.32,33 These pro-apoptotic proteins insert into the mitochondrial outer membrane where they can behave as death channels, allowing the release of apoptogenic proteins from the intermembrane space that are responsible for the activation of the signal cascade that eventually leads to cell death through caspase-9 and -3-dependent proteolysis. It has been described that the intrinsic and extrinsic pathways (receptor-mediated apoptosis) may be interconnected by the activation and subsequent translocation of the protein Bid (a substrate of caspase 8), eventually leading to mitochondrial permeabilization.34 An irreversible loss of mitochondrial function or the release of some mitochondrial death effectors, like apoptosis-inducing factor, may initiate a suicide programme by a caspase-independent pathway. Other cell death effectors released by mitochondria are cytochrome C, second mitochondrial-derived activator of caspase (Smac), and endonuclease G, among others.32
Mitochondria-induced cell death may also be initiated by the development of a sudden increase in the permeability of the mitochondrial inner membrane, the so-called permeability transition (MPT), which allows the passage of ions and solutes with molecular masses up to 1500 Da.35 This phenomenon leads to the dissipation of ΔΨm, uncoupling of oxidative phosphorylation (ATP depletion), and matrix swelling and is triggered by oxidative stress and matrix calcium overload. MPT is modulated by voltage, adenine nucleotide concentration and pH,35 and by pharmacological agents.36,37 The exact molecular nature of MPT remains elusive,38,39 but its contribution to the pathophysiology of multiple cardiac diseases is unequivocal and has been extensively reviewed.39
The connection between mitochondria and endoplasmic reticulum
The endoplasmic reticulum (ER) is a complex, folded membrane network composed of tubules, vesicles, and cisternae expanded within the cytoplasm of eukaryotic cells. It is divided into rough (ribosome-containing) ER, smooth ER, and nuclear membrane, and it fulfils several cellular functions, including synthesis of lipids and steroids, metabolism of carbohydrates, manufacturing and folding of proteins, regulation of calcium concentration, drug detoxification, and attachment of receptors on cell membrane proteins.40 The lumen of the ER contains the highest calcium concentration within the cell, which is due to the activity of calcium ATPases, and a very rich oxidative environment necessary for the formation of protein disulfide bonds and proper protein folding.41 As a result, the ER is particularly sensitive to disturbances in redox regulation and calcium homeostasis, which can promote the unfolded protein response, an activation of a signal transduction event aimed at accelerating degradation of defective proteins.42 Upon persistent noxious stimulus, the ER can initiate cell death via mitochondria-dependent and -independent apoptotic pathways.43 Thus, there is a close interconnection between ER stress and cell survival. The Bcl-2 family proteins, first described by their actions on mitochondrial death pathway, may also interact with ER membranes, and they have been described to modulate ER calcium44 and protect mitochondria through regulation of ER–mitochondria communication.45
Mitochondria also have an important role as intracellular reservoirs of calcium. Interestingly, there is a clear discrepancy between mitochondrial calcium kinetics observed in in vitro preparations and in intact cells. Studies with calcium microelectrodes or with fluorescent markers have established that energized, isolated mitochondria may take up significant amounts of calcium when exposed to high calcium concentrations—in the high micromolar range—through the low-affinity calcium uniporter, without suffering permeability transition.46,47 However, the cytosolic calcium concentration in resting cells is well below this level, and even maximal calcium concentrations reached after physiological stimulation (1–3 µM) appear to be insufficient for mitochondrial calcium uptake, according to the described calcium affinity of the mitochondrial calcium uniporter. The studies from Rizzuto et al.48,49 using chimeric recombinant aequorin targeted to mitochondria established that the mitochondrial calcium concentration increases rapidly and transiently upon the stimulation of ER calcium release (IP3 agonists) in different quiescent cell types, although the speed and amplitude of the rise in mitochondrial calcium levels was not accounted for by the relatively small increases in mean cytosolic calcium. These and other similar experimental approaches showed that mitochondria respond more efficiently to calcium increases in the specific sites located close to the ER calcium channels than to increases in general cytosolic calcium, and that only a fraction of mitochondria are exposed to these microanatomical sites.50 The existence of a physical coupling between ER and mitochondria has been demonstrated in experiments in which limited proteolysis with trypsin or proteinase K significantly modified mitochondrial calcium signalling evoked by IP3 receptor-mediated calcium release.51 Cross-talk between mitochondria and ER has been confirmed in many cell types, including astrocytes,18 hepatocytes,52 and cardiac myocytes53,54 and has been unequivocally established by electron microscopy using fixed samples.55 According to electronic tomography measurements, the distance between the outer mitochondrial membrane and the ER in situ has been calculated to be 10–50 nm.51 These findings agree with the existence of local calcium microdomains that allow privileged signalling between the ER and mitochondria that are responsible for the microheterogeneity of cellular calcium.56 Mitochondria would act as calcium sensors in a discrete number of functional microdomains (in close juxtaposition with the ER) and would allow the diffusion and amplification of the calcium signal following a well-coordinated pattern. This concept has been proved in experiments in which high-speed mitochondrial calcium imaging revealed propagating intramitochondrial calcium waves in intact cells that were susceptible to interruption by overexpression of the mitochondrial fission factor Drp-1.57 A handful of biochemical and functional results have implicated several proteins in sustaining ER–mitochondria coupling.58–61 The preferential communication between ER and mitochondria has emerged as a concept with great biological relevance in a variety of tissues, and more specifically in those cells with a high degree of specialized functions.62,63
The case of cardiac myocytes: communication between mitochondria and sarcoplasmic reticulum
Specific characteristics of mitochondria and sarcoplasmic reticulum in cardiac cells
Cardiac myocytes are the paradigm of highly compartmentalized cells, with a terminally differentiated morphology and extremely specialized functions. Their large size, as well as the uninterrupted mechanical work they exert throughout life, makes cardiac myocytes the most energy-demanding cells in the body. The development of contractile activity is also possible by a precise and perfectly coupled calcium oscillatory movement. These requirements have impacted on the specialization and abundance of cardiac mitochondria and sarcoplasmic reticulum (SR), the two organelles that coordinate energy production and calcium control. Cardiac mitochondria occupy as much as 40% of cell volume and are unique, in that they can be morphologically and biochemically differentiated into two subpopulations: subsarcolemmal mitochondria, located beneath the plasma membrane, and interfibrillar mitochondria, located between myofibrils.64 Morphological differences in the two types of mitochondria are subtle, but respiratory activity and the capacity to retain calcium are significantly higher in interfibrillar mitochondria.65 It has been recently described that connexin 43, a protein whose translocation to mitochondrial membranes has been associated with cardioprotection during ischaemic preconditioning,28,29 is present only in subsarcolemmal mitochondria.66 Similarly, functional decline associated with ageing is predominantly manifested in interfibrillar mitochondria,67 and diabetic interfibrillar mitochondria have been shown to display a greater propensity for undergoing apoptosis when compared with subsarcolemmal mitochondria.68 It can be speculated that biochemical differences between the two mitochondrial subpopulations may be the result of their different interaction with other intracellular organelles.
The SR shares several morphofunctional features with the ER, like the complex tubular structure and cisternae, but they differ in some biochemical characteristics and interactions with other cell structures. In striated skeletal and cardiac cells, the SR coexists with the classical ER-containing ribosomal network. The SR constitutes the main intracellular calcium store in striated muscle and plays a crucial role in excitation–contraction coupling. Owing to its high degree of functional specialization, the vast majority of SR protein content in cardiac cells corresponds to calcium transporters SERCA at the longitudinal region, and calcium release channels (ryanodine receptors, or RyR) at the region juxtaposed to T-tubules.56,69 SERCA belongs to a family of ATP-dependent calcium transporters that maintain intraluminal calcium concentration 3–4 orders of magnitude greater than cytosolic calcium,70 and it is regulated by phospholamban, a 22 kDa protein present in the longitudinal fraction of SR.71 SERCA2 is the only isoform expressed in cardiac cells. RyR mediate the release of calcium from SR to myofibrils during cell contraction. RyR channels are blocked by micromolar concentrations of ryanodine and activated by millimolar concentrations of caffeine, and they are the main components responsible for calcium-induced calcium release, a cycle of positive feedback, in cardiac cells.72 RyR2 is the only isoform expressed in myocardial tissue. RyR have a complex regulatory mechanism, and several factors have been described to modify their open probability apart of calcium, like oxidation of critical residues, phophorylation/dephosphorylation events, and protein interactions (mainly with calmodulin, FKBP12.6, and calsequestrin).3,73,74 Moreover, SR calcium storage capacity is regulated by a series of intraluminal, low-affinity calcium-binding proteins. During calcium transients, only 40–60% of total SR calcium is released.75,76 Calsequestrin, the most abundant intraluminal protein, acts as a calcium sensor capable of inhibiting RyR activity at low calcium concentrations and terminating SR calcium release during normal cell contraction.77 Several other less-represented SR proteins also participate in SR calcium regulation.78 Efficient delivery of calcium and ATP to the sarcomere is possible because of an intimate structural interaction between T-tubules, RyR channels, myofilaments, and mitochondria. Calcium uptake by mitochondria located in close proximity to SR calcium release channels participates in the modulation of the calcium transient and contraction.79
Functional microdomains between mitochondria and SR
In cardiac myocytes, mitochondria form a dynamic and continuous network surrounding myofibrils and the SR network, within which the inter-organelle distance has been estimated to be 10–50 nm.51 SR–mitochondria communication occurs through anatomical and functional cell microdomains and is in part responsible for the heterogeneous distribution of calcium across the cytosol of cardiac cells. It has been demonstrated that mitochondrial calcium uptake is more dependent on high cytosolic Ca2+ microdomains around the contact sites with SR than on the cytosolic Ca2+ concentration.54 Although it is clear that the interplay between SR and mitochondria allows a rapid exchange of molecules between both organelles,50 the mechanism of such diffusion is less well understood. According to the recent evidence, SR–mitochondrial communication is supported by a physical coupling resistant to purification procedures that cause demolition of cytoskeletal structure.50 By means of conventional transmission electron microscopy as well as electron tomography, it has been possible to visualize tethering structures (electron-dense bridges) between the terminal cisternae of the SR and mitochondria80 (Figure 1). In any case, the mitochondrial calcium uniporter is exposed to local high calcium concentrations, since RyR-mediated calcium propagation to mitochondria persists even after cytosolic calcium chelation with BAPTA in permeabilized cardiac myocytes.81
Cross-talk between mitochondria and SR is not only a determinant of calcium handling but for a proper matching of energy supply and demand and regulation of mitochondrial respiration.82,83 More recently, the concept of ‘ROS microdomains’ has been proposed to describe a spatially restricted ROS generation with different physiological effects, depending on their localization.84 Because mitochondria are the major producers of superoxide, it can be speculated that mitochondrial ROS production may specifically impact SR function and calcium microdomains. In fact, both SERCA pumps and RyR channels contain multiple, potentially redox-sensitive cysteine residues, and cysteine thiol oxidation has been described to increase RyR channel activity.3 Therefore, it is not surprising that photostimulation of mitochondrial ROS production is able to induce a transient increase in SR calcium sparks in rat cardiac myocytes.85 Remarkably, the relationship between ROS and calcium at the locally restricted cell domains may take place in both directions. A recent report86 has shown that local increases in cytosolic calcium concentration, initiated by calcium release from the SR, increased ROS production and favoured MPT. Thus, the existence of a mutual relationship between calcium and ROS signalling is favoured by the SR–mitochondria interface and probably represents one of the most important signalling mechanisms in cardiac cells.
Mitochondria and SR in cardiovascular pathophysiology
The interplay between SR and mitochondria has some genuine characteristics that go beyond the individual role of each organelle. Functional consequences of the close anatomical interaction between SR and mitochondria are increasingly recognized as playing important roles in the pathophysiology of several conditions. Truncation or modification of the physiological behaviour of one of the components of these microdomains may impact on the other, amplify cell injury, and even induce necrotic or apoptotic death.87,88 Defects in the intercommunication between SR and mitochondria make cardiac myocytes particularly vulnerable to problems related to ATP availability, sarcomeric calcium, and oxidative stress.
SR–mitochondria communication may amplify ischaemia–repefusion injury
Cell death secondary to transient myocardial ischaemia occurs mainly within the first minutes of restoration of blood flow in the form of necrosis. The mechanisms of reperfusion-induced cardiomyocyte death have been reviewed elsewhere.89 Although they have not been completely elucidated, calcium overload and altered calcium handling play a prominent role by inducing calpain-mediated proteolysis, hyperactivation of the contractile machinery, and MPT. The pathological hallmark of reperfusion necrosis is the presence of contraction bands, reflecting cardiomyocyte hypercontracture.90,91 Energy-dependent hypercontracture92 contributes to cell death, as demonstrated by studies in which pharmacological inhibition of contractility at the onset of reperfusion reduced infarct size.93–95 Previous studies have indicated that hypercontracture can be triggered by SR calcium cycling96,97 induced by restoration of ATP in the presence of cytosolic calcium overload. The reactivation of SERCA causes SR calcium uptake beyond maximal SR storage capacity, resulting in subsequent calcium release through RyR and an increase in the concentration of cytosolic calcium, reinitiating the process. This abnormal SR behaviour gives rise to cytosolic calcium oscillations that propagate as calcium waves and may induce arrhythmias and myofibrillar hypercontracture.96,97 The prevention or attenuation of calcium oscillations by drugs interfering with SR calcium uptake or release96,98 or reducing cytosolic calcium overload99,100 has been proved to be effective strategies to protect against hypercontracture and cell death during reperfusion.
The role of mitochondria in myocardial reperfusion injury has been the object of extensive research and has been reviewed in detail previously.101 Oxidative stress, mitochondrial Ca2+ overload, and low ATP concentration—circumstances that concur upon myocardial reperfusion after prolonged ischaemia—promote MPT in different experimental models.101–103 Pharmacological and genetic inhibition of MPT effectively reduced infarct size in several experimental models,36,37,104 and two proof-of-concept clinical trials have demonstrated that targeting MPT with postconditioning105 or cyclosporine A106 at the moment of angioplasty improves contractile function and reduces cell death in patients with acute myocardial infarction. However, hypercontracture is an energy-dependent response, whereas MPT causes energy dissipation. One interesting possibility is the coexistence within the same cell of an intact subpopulation of mitochondria, capable of sustaining ATP synthesis, with more severely damaged mitochondria undergoing membrane permeabilization. The anatomo-functional interplay between the SR and mitochondria predicts, in fact, a different tolerance to ischaemic damage of subsarcolemmal and interfibrillar mitochondria, since only the latter subpopulation is in close contact with the SR. The coexistence of damaged and intact mitochondria within the same cell during reperfusion has been tested in a model of laser-induced mitochondrial permeabilization in calcium-overloaded cadiomyocytes.107 In this study, experimental induction of MPT impaired cytosolic calcium overload, favoured SR-driven calcium oscillations, and eventually led to hypercontracture, supporting the existence of heterogeneously damaged mitochondria within the same cell. The relationship between SR-driven calcium oscillations and MPT appears to be bidirectional, and SR calcium cycling may increase the susceptibility for MPT in a fraction of mitochondria located in areas of close contact with the SR. In a recent study,108 pharmacological blockade of SR calcium load with thapsigargin/ryanodine slowed mitochondrial calcium uptake in intact cells but had no effect in isolated mitochondria where the contribution of SR is negligible. The inhibition of SR calcium uptake and release also reduced MPT, hypercontracture, and cell death during reperfusion, and this protective effect was abrogated when anatomical interaction between the SR and mitochondria was partially disrupted with colchicine108 (Figure 2). The prevention of cytosolic calcium oscillations did not reduce the total cytosolic calcium concentration during the first minutes of reperfusion, supporting the hypothesis that there is a component of cell death triggered by a local mechanism not related to total cellular calcium load.108Figure 3 illustrates the concept of the role of SR–mitochondria microdomains in reperfusion injury.
Arrhythmias may be the consequence of a defective SR–mitochondria interplay
Imbalanced SR–mitochondria communication may have important functional consequences in cardiac arrhythmias. Mitochondrial calcium is the main regulatory factor for the orchestration of mitochondrial energetics with variations in cardiac workload and excitation–contraction coupling.79,109,110 Some studies have linked mitochondrial dysfunction to certain types of cardiac arrhythmias and sudden death.111 Atrial myocyte energetics has been reported to be perturbed in patients and animal models of atrial fibrillation,112 and mitochondrial depolarization during reperfusion—secondary to calcium uptake—could facilitate mitochondrial network disruption and the occurrence of ventricular fibrillation.113 Whether mitochondrial dysfunction is the cause or consequence of SR dysfunction remains to be elucidated. It is postulated that calcium-dependent arrhythmias are generally associated with the instability of SR calcium release, whereas heart failure is related to a reduction in systolic SR calcium release.114 The pathophysiological mechanism of SR dysfunction is extremely complex and multiple molecular disturbances, like abnormal RyR2 behaviour with increased SR calcium leak115 or failure of RyR2 gating,116 have been described. Moreover, conditions causing an increase in the cytosolic calcium concentration, i.e. myocardial ischaemia/reperfusion, may lead to the generation of spontaneous SR calcium waves, which can give rise to sustained tachyarrhythmias117 and promote excessive activation of the contractile machinery, leading to hypercontracture and cell death.96,97 The SR has been pointed out as the major contributor of the contractile dysfunction in human atrial cardiomyocytes obtained from patients with atrial fibrillation by a mechanism involving hyperphosphorylation of the RyR2.118 This effect could be responsible for a continuous SR calcium leak and arrhythmogenesis. Alternatively, the increase in mitochondrial ROS production described in human atrial fibrillation119 could be in part responsible for SR dysfunction. Computer simulation studies integrating the role of excitation–contraction coupling and mitochondrial bioenergetics showed that mitochondrial ROS production triggers ΔΨm oscillations and shortens the duration of the cell action potential.120 The possibility that altered SR function may locally modify mitochondrial energetics deserves further investigation.
Disruption of mitochondria–SR interaction in the failing heart
Advanced heart failure has been consistently associated with a progressive decline in mitochondrial respiratory activity,121 reduced ATP generation,122 and mitochondrial structural abnormalities.123 These mitochondrial alterations, together with reduced amplitude of SR-driven calcium transients,124,125 are the most prominent disturbances described to contribute to cardiac contractile dysfunction. A large variety of defects in the activity of individual components of the mitochondrial respiratory chain have been proposed to participate in the pathogenesis of heart failure.126–128 It is controversial whether the reduction in the respiratory rates may be attributed to a mere decrease in mitochondrial function or to a reduced number of mitochondria.129 Studies from Rosca et al.121,130 propose that the pronounced variability of mitochondrial electron transport defects reported thus far in severe acquired cardiomyopathies may be explained by a sequential mechanistic pathway in which defects in supramolecular assembly (respirasomes)—rather than in the individual components of oxidative phosphorylation—are the primary event responsible for the progression of heart failure. Recently, it has also been suggested that changes in mitochondrial protein dynamics can contribute to the cell loss and the progression of cardiac failure.131 Moreover, in some specific conditions like heart failure secondary to diabetic cardiomyopathy, interfibrillar and subsarcolemmal mitochondria are selectively affected.68,132 Mitochondria-induced apoptotic death of cardiac myocytes has been described to play a role in the progression of cardiac dysfunction, since it is persistently activated in patients with decompensated heart failure.133
Because in cardiac cells, the matching of energy supply and demand is strongly dependent on local, metabolically active microdomains, the impairment of the local ATP/ADP ratio is one of the molecular disturbances described in failing myocardium, probably developing as a consequence of a cytoarchitectural perturbation.134,135 Indeed, heart failure is associated with structural remodelling of various subcellular organelles,136 and it has been described that local energy transfer between SR and mitochondria is affected in mice with genetically altered subcellular organization.137 Mouse hearts deficient in muscle LIM protein exhibit a defective nucleotide channelling from mitochondria to SR, despite having normal mitochondrial oxidative capacity, an effect that likely accounts for the energetic and contractile dysfunction.137 Therefore, there is a tight relationship between the disorganization of several cellular structural components and the impairment of SR–mitochondria intercommunication, with concomitant deterioration in the excitation–contraction coupling.135,137 Although the specific role of SR–mitochondria interplay in the pathogenesis of the different cardiomyopathies (dilated vs. hypertrophic) has not been elucidated, subcellular microanatomical remodelling seems to be dependent on the differences in cardiac load and cardiac failure stage.135
Disarrangement between SR and mitochondria is not the only mechanism responsible for the defective biochemical interaction among both organelles. Alterations of structural or regulatory proteins on any side of the functional unit may have consequences on its counterpart. Thus, the modification of the expression of the uncoupling protein 2 not only has an intrinsic mitochondrial effect on ΔΨm and ROS production but also exerts a deleterious influence on excitation–contraction coupling and calcium handling.138 Conversely, calcium/calmodulin-dependent protein kinase II, a protein that plays a role in regulating SR calcium release, may reduce ΔΨm and increase mitochondrial susceptibility to undergoing membrane permeabilization by a mechanism related to local rise in cytosolic calcium and stimulation of mitochondrial ROS generation.86 Overall, these studies emphasize the critical role of the disruption of functional units, particularly those established by SR and mitochondria interplay, in the progression of heart failure.
Role of mitochondria–SR microdomains in diabetes
Several lines of evidence support the notion that mitochondrial defects play a critical role in the progression of diabetes.139,140 Similarly, a number of studies have found altered SR function with disturbed calcium handling in the cardiomyopathic phenotype associated with the diabetic state.141,142 Cardiac myocytes submitted to hyperglycaemic conditions are exposed to increased oxidative stress and mitochondrial calcium overload, which can lead to a reduction in mitochondrial respiration and increased susceptibility to undergoing MPT.143 Mitochondrial perturbations have been attributed to impaired myocardial insulin signalling144 or to a fragmentation of the mitochondrial network.145 Defective nitric oxide production might also deteriorate mitochondrial biogenesis in the metabolic syndrome.146 Indirect evidence suggests an interaction between SR and mitochondrial disturbances in the progression of diabetes. Reduced insulin action promoted a rapid decline in mitochondrial fatty acid oxidative capacity.147 Interestingly, the expression levels of SERCA2 were concomitantly reduced.147 A different study148 indicated that chronic exposure to high fatty acid levels adversely affected mitochondrial function (reduced ΔΨm and stimulated ROS production), which, in turn, was associated with reduced SR calcium content and depressed cytosolic calcium transients. It cannot be ruled out that SR (or ER) stress mediates mitochondrial damage by a cell signalling mechanism rather than by a direct local effect. Miki et al.149 described a reduced threshold for MPT associated with ER stress in the myocardium of type 2 diabetic rats by an impaired phospho-GSK3β-mediated suppression of MPT, an effect that may underlie the failure of cytoprotective strategies against ischaemia–reperfusion injury in diabetic hearts. Dysregulation in the formation of ROS microdomains could play a role in the pathophysiology of diabetes due to the existence of a tight interplay between ROS and calcium, but, to date, the specific contribution of SR–mitochondria interaction in the aetiopathogenesis of diabetic disease has not yet been explicitly investigated.
Might ageing be a disease of SR–mitochondria communication?
Ageing is a complex, multifactorial process that is far from being deciphered. It is usually linked to degenerative diseases, including cardiac diseases, but the nature of this link remains speculative.150 An overwhelming amount of experimental data links mitochondrial dysfunction with cell senescence and lifespan.151 The mitochondrial theory of ageing assumes three main mediators of cell damage: increased ROS accumulation, progressive mitochondrial DNA damage, and insidious failure of respiratory complexes.151,152 However, to our knowledge, the possibility that different populations of mitochondria age at different rates depending on their interrelation with the SR, orchestrating cell senescence, has not been investigated thus far. An intriguing observation is that subsarcolemmal and interfibrillar mitochondria experience different functional decline during ageing, with a selective reduction in oxidative phosphorylation affecting interfibrillar mitochondria.153 The capacity to retain calcium before undergoing membrane permeabilization is also specifically reduced in the interfibrillar mitochondria from myocardium of old rats.154 Remarkably, interfibrillar mitochondria are exposed to continuous SR calcium oscillations throughout the life in the beating heart. This is probably the biological reason for their higher respiratory capacity and a more efficient calcium uptake system.65 Because the SR–mitochondria interaction is bidirectional, the possibility exists that progressive mitochondrial decline in oxidative phosphorylation observed in interfibrillar mitochondria deteriorates the closely juxtaposed, energy-dependent SR, providing the pathophysiological basis for the progression of age-related degenerative diseases in which calcium homeostasis is lost, i.e. heart failure. Ageing has been shown to account for an increased sensitivity to ischaemia–repefusion injury mediated in part by a higher diastolic calcium concentration, although the mechanisms leading to this persistent loss of calcium control have not been identified.155 The activity of SR calcium transporters is very sensitive to redox state,3 and functional units regulating local SR calcium release and the calcium-induced-calcium-release response in skeletal and cardiac cells have been shown to lose their plasticity during ageing.156 However, what the molecular machinery that induces such alterations might be is not clear at present, nor is the cause–consequence relationship between them, and the role of the SR–mitochondria interaction in the cellular functional decline associated with ageing is still an unexplored field of research.
The concept that mitochondria and SR behave in many respects as a functional unit has potentially important implications. First, it may allow a better understanding of the pathophysiology and evolution of cardiovascular diseases and their modulation by ageing. Secondly, and most importantly, it may help to identify and develop new therapeutic strategies. It should allow the protection of mitochondria by targeting SR molecules or treatment of SR dysfunction by mitochondrial interventions. Finally, it opens a new field of research on the potential value of strategies aimed at modifying SR–mitochondria communication itself.
Conflict of interest: none declared.
Partially supported by RECAVA-RETICS (RD060014/0025, ISCIII, Spanish Ministry of Science), FIS (PS09/02034, Spanish Ministry of Science), and SAF (2008-03067, Spanish Ministry of Education).