Essays in Biochemistry

Roles of mitochondria in human disease

Michael R. Duchen, Gyorgy Szabadkai


The chapters throughout this volume illustrate the many contributions of mitochondria to the maintenance of normal cell and tissue function, experienced as the health of the individual. Mitochondria are essential for maintaining aspects of physiology as fundamental as cellular energy balance, the modulation of calcium signalling, in defining cellular redox balance, and they house significant biosynthetic pathways. Mitochondrial numbers and volume within cells are regulated and have an impact on their functional roles, while, especially in the CNS (central nervous system), mitochondrial trafficking is critical to ensure the cellular distribution and strategic localization of mitochondria, presumably driven by local energy demand. Maintenance of a healthy mitochondrial population involves a complex system of quality control, involving degrading misfolded proteins, while damaged mitochondria are renewed by fusion or removed by autophagy. It seems evident that mechanisms that impair any of these processes will impair mitochondrial function and cell signalling pathways, leading to disordered cell function which manifests as disease. As gatekeepers of cell life and cell death, mitochondria regulate both apoptotic and necrotic cell death, and so at its most extreme, disturbances involving these pathways may trigger untimely cell death. Conversely, the lack of appropriate cell death can lead to inappropriate tissue growth and development of tumours, which are also characterized by altered mitochondrial metabolism. The centrality of mitochondrial dysfunction to a surprisingly wide range of major human diseases is slowly becoming recognized, bringing with it the prospect of novel therapeutic approaches to treat a multitude of unpleasant and pervasive diseases.


Mitochondria are the primary providers of ATP in most mammalian cells, they regulate both necrotic and apoptotic cell death pathways, they are intimately involved with the co-ordination of cellular Ca2+ signalling and they both generate and are targets of free radical species that modulate many aspects of cell physiology (see Figure 1). It therefore inevitably follows that disordered mitochondrial function will lead to altered cellular physiology and that this will manifest in disease. In this chapter, we will try to identify some of the major cellular mechanisms which contribute to the pathophysiology of mitochondrial-related disease and to explore the functional consequences of such changes for cells and tissues so that we can better understand the roles of mitochondrial dysfunction as a determinant of the phenotypic manifestations of disease.

Figure 1 A cartoon showing major pathways for cell death and altered mitochondrial calcium signalling in Parkinson-related PINK1-knockout cells

The tricarboxylic acid cycle (TCA cycle) maintains the reduced state of the intermediates NADH (and FADH2) which supply electrons to the complexes of the respiratory chain, which transfers electrons to oxygen and translocates protons across the IMM (inner mitochondrial membrane), so establishing the potential (Δψm). The potential provides the energy that drives ATP synthesis at complex V and that also drives calcium influx through the uniporter. Free radical species may be generated in the respiratory chain primarily at complex I and complex III. Under pathological conditions, a combination of increased free radical generation and resultant oxidative stress combined with high intramitochondrial calcium may cause opening of the mPTP (mitochondrial permeability transition pore). This will collapse Δψm and may also cause mitochondrial swelling, release of cytochrome c and AIF (apoptosis inducing factor) from the intermembrane space. Cytochrome c activates the caspase cascade and AIF translocates to the nucleus, together driving pathways for apoptotic cell death. Calcium enters mitochondria when cytosolic calcium is raised, and it is removed by the sodium/calcium exchanger (mNCX). The inset traces show how, in mutations of PINK1 associated with PD, calcium removal by the mNCX is dramatically slowed following a stimulus to raise calcium (arrow), leading to mitochondrial calcium overload, mPTP opening and a loss of Δψm (as illustrated). WT, wild-type.

At its extreme, we tend to focus on roles of mitochondria in cell death as the major role of mitochondria in disease, and this is without doubt hugely important. However, in this chapter, we will also take a far broader view of the roles of mitochondrial (dys)function in disease. This will include the down-regulation of mitochondrial OXPHOS (oxidative phosphorylation) which accompanies the shift towards a more glycolytic phenotype of neoplastic cells defined as the Warburg shift, indicating that mitochondria play critical roles in many cancers, not only through the suppression of apoptotic cell death, but also through the modulation of mitochondrial metabolism. We will consider changes in mitochondrial trafficking that lead to neuropathic disorders, changes in mitochondrial fission and fusion that seem so important for mitochondrial ‘quality control’, changes in mitochondrial mass through changes in turnover, i.e. either mitochondria-specific macroautophagy (mitophagy) or altered biogenesis. We will also consider further mechanisms by which altered mitochondrial function may contribute to disease, which seem to be recapitulated repeatedly in many apparently diverse diseases. One repeated theme in mitochondrial pathogenesis is the mitochondrial generation of reactive free radical species such as superoxide or peroxide [often termed ROS (reactive oxygen species)] which may, in turn, impact on Ca2+ signalling, on cell death signalling pathways and indeed on mitochondrial turnover. In many instances, multiple mechanisms may create a cascade of effects that all contribute to disease. Changes in mitochondrial Ca2+ handling may alter the dynamics of cellular Ca2+ signals that are themselves critical in defining cell function and so will lead to impaired tissue function if disturbed, while changes in intramitochondrial [Ca2+] may sensitize mitochondria to injury. It should be clear then that the potential roles of mitochondria in disease, and the array of mechanisms through which altered mitochondrial function may contribute to pathogenesis, are many and varied, and that to understand all of these, we need to understand the fundamental phenomena involved: the mechanisms of biogenesis, trafficking, autophagy, free radical generation and calcium handling, and mitochondria-mediated cell death, most of which are covered elsewhere in this volume.

In order to approach this very broad topic, it is perhaps simplest to consider mitochondrial function in terms of inherited (primary) and acquired (secondary) defects. The first includes disorders in which alterations in proteins encoded by either the mitochondrial or nuclear genome impact on mitochondrial function and so lead to disease. The second category includes disorders in which changes in cell physiology, changes in mitochondrial biogenesis or autophagy, changes in extramitochondrial signals, such as impaired provision of substrate or oxygen, exposure to cytokines, changes in intracellular Ca2+, oxidative stress or exposure to xenobiotics cause mitochondrial dysfunction and so contribute to the progression or outcome of disease. However, many diseases show a multifactorial aetiology and it is difficult to distinguish between the primary and secondary origin of the accompanying mitochondrial dysfunction. We have therefore organized the chapter according to pathomechanisms where they are known and to disease manifestations in those examples which seem more multifactorial.

Defects in mtDNA (mitochondrial DNA) and disease

Mitochondria house their own very much reduced gene pool as mtDNA, packaged into nucleoids [1]. This DNA encodes just 13 key proteins of OXPHOS, along with rRNAs (ribosomal RNAs; 12S and 16S rRNA) and 22 tRNAs (transfer RNAs) required for the synthesis of these proteins [24]. Inevitably it follows that mutations will arise, and those which are not lethal will give rise to disease. mtDNA is transmitted through a non-Mendelian maternal pathway and so mitochondrial genetic disease is characteristically inherited through the maternal family line [5]. The genetics are even more complicated, as the majority of mtDNA genetic diseases are heteroplasmic: subjects carry both normal and mutant mtDNA [6,7]. It is assumed that the burden of that mutant load, its partitioning between tissues and the interaction between the nuclear and mitochondrial genomes will define the actual pathological phenotype [8]. However, except for some cases, the exact relationship between genotype and phenotype remains obscure. For example, there is no explanation why one mutation of a complex I subunit gives rise to a disease that causes primarily only optic atrophy developing in adolescence or early adult life {LHON (Leber’s Hereditary Optic neuropathy); for a review see [9]}, while another mutation in the same complex may give rise to a severe myopathy, neuropathy and lactic acidosis in early childhood (e.g. see [10]). Pathogenic mechanisms in these diseases also remain poorly understood; most of the diseases affect mostly the CNS (central nervous system) and muscle and this is usually explained by the high dependence of these tissues on OXPHOS, but it is not at all clear that the manifestations of disease are direct consequences of impaired OXPHOS (i.e. in impaired provision of ATP) or whether there are more subtle alterations, for example in free radical or calcium signalling (see below). These issues will be covered in more detail in other chapters in this volume and we simply flag these here for completeness. Recently, several mutations have been detected in numerous cancer tissues, most probably reflecting the anti-apoptotic effect of the mild impairment of OXPHOS (see below).

Defects in nuclear-encoded mitochondrial-related proteins and disease

While mtDNA-encoded proteins are important and mutations at many sites of mtDNA have been identified in association with disease, the vast majority of mitochondrial proteins are encoded in nuclear DNA [1113]. Indeed, the tissue- or cell-specific differences in mitochondrial structure and function that are long familiar to histologists and electron microscopists must largely reflect differences in the palette of nuclear-encoded, mitochondrial-targeted proteins expressed by specific cell types. Once again, it seems inevitable that diseases associated with mutations of impaired function of these proteins will be identified. The number of nuclear-encoded mitochondrial-directed proteins associated with human disease is rapidly growing (e.g. see and [14,15]). Most of these genes encode enzymes of mitochondrial respiratory chain complexes, structural proteins required for the assembly of these complexes, mitochondrial transporters and a number of proteins involved in various aspects of quality control of the organelle, as well as its individual protein components. These include structural proteins and large mechanochemical enzymes (GTPases) that control mitochondrial fission and fusion [16,17], chaperones responsible for correct folding of soluble and membrane proteins, and proteases clearing non-functional, damaged proteins, all involved in various ways in maintaining the integrity of the mitochondrial population (for a recent review, see [18]). It is notable that by far the majority of clinical presentations of the various diseases associated with mutations of these proteins involve disorders of the CNS, emphasizing the importance of mitochondrial function in this tissue, although we still understand little of the mechanisms by which changes in these specific functions cause the specific and varied disease phenotypes, nor indeed why these can differ so radically with different mitochondrial mutations.

Mitochondrial ‘quality control’ and disease: proteases, fission, fusion and turnover

Of the mitochondrial proteins that are involved in mitochondrial quality control, mutations in several have now been associated unambiguously with human disease, in all cases primarily affecting the CNS. These include a number of proteins associated with hereditary (and even some sporadic) forms of PD (Parkinson’s disease), hereditary optic atrophy and motor neuron degeneration. Mitochondrial integrity is maintained partly through the activity of chaperones and intramitochondrial proteases that either support protein folding or that remove misfolded proteins [1922]. These include the Lon protease, which seems to selectively remove oxidatively damaged proteins [23]. Accumulation of oxidatively damaged mitochondrial proteins with age has been attributed, in part, to declining activity of Lon with age, and deletion of Lon from mitochondria increases oxidative cell death, decreases mitochondrial mass and promotes a more glycolytic metabolic phenotype [24]. Another major protein involved in regulation of mitochondrial protein folding is the protease known as paraplegin. Mutations of paraplegin give rise to the motor neuron degenerative disease, hereditary spastic paraplegia [25,26]. This disease is characterized by the degeneration of cortical motor neurons, which begins in their very long axons, suggesting that mitochondrial function at axonal terminals may be critically damaged by accumulation of misfolded proteins [27]. Interestingly, mutations in a mitochondrial matrix chaperone Hsp60 (heat-shock protein 60), also required to maintain correct folding of mitochondrial proteins, also cause an autosomal dominant form of hereditary spastic paraplegia [28], consistent with the suggestion that mitochondrial quality control is critical for maintenance of axons. What remains mysterious, a word we have perhaps already used too often but which remains true, is why global defects in these proteins should give such a very specific phenotypic defect: why do defects in such fundamentally important processes not cause cardiac or liver defects, or even defects in other parts of the CNS? In contrast, another fundamental mitochondrial chaperone, grp75 (also known as mthsp70), which mediates folding as part of the protein import machinery [29] and has been shown to play a role in mitochondrial oxidative damage [30] and cellular senescence [31], is associated with a much broader variety of diseases, ranging from inflammatory and immune disease to cancer [32], in addition to neurodegenerative disease [33].

The dynamic nature of the mitochondrial network also seems to play an important role in mitochondrial quality control. Damaged mitochondria can be either ‘repaired’ or removed by fusion with neighbouring intact mitochondria [34,35] or by autophagy, which removes the damaged organelles while avoiding the potential catastrophe of the release of pro-apoptotic proteins [36,37]. The difference between these two fates seems to depend on the degree of damage, as severely damaged and fragmented mitochondria will not undergo fusion, but may instead be destroyed. Mitochondria constantly undergo cycles of fusion and fission, a process which is mediated by several highly conserved proteins [34,38,39]. At least four dynamin-related GTPases mediate fission and fusion, a process which determines the shape and extent of the mitochondrial network. Fusion is driven by the mitofusins MFN1 and MFN2 which localize to the OMM (outer mitochondrial membrane) and interact with transport structures such as microtubules. The IMM (inner mitochondrial membrane) protein OPA1 (optic atrophy 1), controls mitochondrial fusion, while fission is triggered by DRP1 (dynamin-related protein 1). Fusion seems to promote intermingling of the matrix contents of intact and dysfunctional mitochondria [40]. Thus replacement of damaged mtDNA may contribute to the integrity of the cellular mitochondrial population. In fusion-deficient fibroblasts lacking OPA1 or the mitofusins, mitochondria showed both a loss of mtDNA nucleoids and impaired mitochondrial respiration, suggesting an accumulation of damaged mitochondria [5,41]. Mutations in the mitochondrial fusion gene Mfn2 cause the human neurodegenerative disease Charcot–Marie–Tooth, a sensory neuropathy [42]. Deletion of Mfn2 from the cerebellum caused impaired dendritic outgrowth, spine formation and cell survival of Purkinje cells, in which mitochondrial distribution, ultrastructure and electron transport chain activity were all abnormal. Thus exchange of mitochondrial contents plays an important role in mitochondrial function and distribution in neurons.

Fragmented and damaged mitochondria can be removed by a mitochondrial-specific form of autophagy; the term mitophagy was recently coined [36]. Studies in yeast have identified ∼30 proteins (Atg) involved in autophagy, many of which are conserved in higher eukaroytes [43]. Deletion of yeast ATG genes causes phenotypes which reflect altered mitochondrial function, including growth defects on non-fermentable growth medium, suggesting a requirement for autophagy in mitochondrial maintenance [44]. Two mitochondrial-specific proteins, termed Uth1 and Aup1, are not part of the general autophagic machinery and have been linked specifically to mitophagy: Uth1 localizes to the mitochondrial outer membrane and Aup1, a phosphatase, to the intermembrane space [45,46].

Accumulating evidence suggests that autophagy is initiated by increased mitochondrial ROS generation [47,48]. This seems to be mediated by the redox regulation of a protein known as Atg4, an essential cysteine protease in the autophagic pathway [49]. Formation of autophagosomes requires the Atg4-mediated cleavage of Atg8 and its conjugation to phosphatidylethanolamine, a lipid constituent of autophagosomal membranes [43]. Antioxidant treatment suppresses lipidation of Atg8 and autophagosome formation, suggesting that lipid conjugation is a direct consequence of ROS production. Increased chronic ROS generation by damaged mitochondria has also been shown to promote autophagy and so reduce the mitochondrial volume fraction in cells in which the protein IF1 has been knocked down [47]. In this model, autophagosome formation and mitochondrial volume were restored by antioxidant treatment. Interestingly, endophilin B1 (also known as Bif-1), first described as a factor in the mitochondrial outer membrane necessary for the maintenance of the organelle structure [50], was later shown to be part of the protein complex forming the early autophagosome [51]. Down-regulation or missense mutations of this protein have been associated with tumour formation [52], again underlying the importance of mitochondrial turnover for cell growth and proliferation.

Mitochondria and neurodegenerative disease

Mitochondrial dysfunction has been implicated in almost all of the major neurodegenerative and neuroinflammatory diseases, including AD (Alzheimer’s disease), PD, Huntington’s disease, Friedreich’s ataxia, motor neuron disease [ALS (amyotrophic lateral sclerosis)] and multiple sclerosis. There are others that have been mentioned already, Charcot–Marie–Tooth and other neuropathies including diabetic sensory and autonomic neuropathy, hereditary spastic paraplegia and others, which are all clearly and directly associated with impaired mitochondrial function. Friedreich’s ataxia is caused by a mutation of frataxin, a mitochondrial protein required for normal iron homoeostasis in mitochondria; iron is critical as a component of several parts of the respiratory chain. Yet again, the mystery is why this defect should be specific for the spinal sensory and motor tracts which degenerate, why these axons become demyelinated, and why other tissues do not show major defects (for a review see [53]). Evidence for mitochondrial contribution to pathophysiology of PD is fairly strong, and we will elaborate on this below. In AD and ALS the story is far more complex and, while there seem to be disorders of mitochondrial function associated with both diseases, it is hard to disentangle which of the observed changes are primary and which secondary in the progression of disease. Even if the changes are secondary, it does not mean that they are not potentially interesting as if they contribute to disease progression, they still may represent valuable therapeutic targets (see [5456]), especially given the lack of satisfactory treatment strategies for these dreadful diseases.

PD is a common age-associated disease, characterized by degeneration of dopaminergic neurons in the substantia nigra which leads to a very characteristic movement disorder. Most patients with PD have sporadic disease of unknown cause, but in a small proportion of patients the disease is familial. In these families, and to some extent in sporadic disease, the disorder has been linked to a number of mutations of proteins which are mostly associated with mitochondrial function in ways that are not yet entirely clear. Mitochondrial dysfunction appears to play a prevalent role in the pathogenesis of the disease [57,58]. Defects of mitochondrial complex I have been specifically and repeatedly associated with PD. Chronic treatment of mice with rotenone, an inhibitor of complex I, where systemic treatment might be expected to cause very general and nasty toxic effects, remarkably leads to a parkinsonian syndrome (one wonders why not a cardiomyopathy for example?). Gene defects in a number of proteins have been shown to associate with PD, including α-synuclein, Parkin, DJ-1, PINK1 (PTEN-induced putative kinase 1, also known as PARK6) and LRRK2 (leucin-rich repeat kinase 2) and HTRA2/OMI [59]. All of these proteins are in some way associated with mitochondria and apparently form part of a cascade of related interacting proteins. Heterozygous missense mutations in HTRA2/OMI have been found in sporadic cases of PD. HtrA2, homologous with bacterial Deg proteases, localizes to the mitochondrial intermembrane space where it is thought to protect against mitochondrial stress. HtrA2 may act to degrade misfolded polypeptides in the mitochondrial intermembrane space or, in analogy to bacterial DegS, be part of an adaptive stress signalling cascade, and HtrA2-deficient mice show neurodegeneration and Parkinson-like phenotypes [60,61].

HtrA2 associates with the mitochondrial-localized protein PINK1, a serine/threonine kinase, mutations of which have also been found associated with PD [62]. PINK1 is required for phosphorylation of HtrA2, which increases its proteolytic activity in vitro. PINK1-mediated phosphorylation has been demonstrated for TRAP1 [TNF (tumour necrosis factor) receptor-associated protein 1], a putative molecular chaperone with significant homology with the HSP90AA1 family [62a]. Overexpression of PINK1 protects cells from apoptosis induced by oxidative stress, while PINK1-knockout cells are more vulnerable to apoptosis inducers [63], suggesting that PINK1 and TRAP1 are part of an anti-apoptotic signalling cascade. We found recently that mitochondrial Ca2+ handling was defective in cells in which PINK1 is knocked down (Figure 1). Thus, following a calcium load, mitochondria were not able to restore intramitochondrial Ca2+ concentration ([Ca2+]m). This was attributed to a defect in the mitochondrial sodium/calcium exchanger which is required for the rapid restoration of [Ca2+]m. As a result, mitochondria quickly became calcium overloaded and cells showed a much reduced threshold for mPTP (mitochondrial permeability transition pore; see below and Figure 1) opening [64], demonstrating what may be a common paradigm – the interaction of a genetic defect with environmental or cell signalling events as a basis for pathophysiology (see below). PINK1 mutations have also been associated with altered mitochondrial fission and fusion when knocked out in Drosophila, while the associated PD-related protein PARKIN has also been related to the targeting of mitochondria for autophagy [65]. It remains a tantalizingly complex field and it is difficult to understand now exactly how these various strands will eventually fit together, but the whole complex tale places mitochondria centre stage in the pathogenesis of a common and serious neurodegenerative disorder that causes havoc in far too many lives.

Mitochondrial dysfunction in aging

The role of accumulating mitochondrial defects as a cause of aging remains highly controversial. There is a widespread belief that mitochondrial-generated oxidative stress and accumulated gene defects may itself cause aging. A recent TV documentary likened accumulated oxidative stress in aging to the oxidation involved in a rusting motor car. It is not at all clear how valid such analogies really are. Through an ingenious strategy, two groups have generated mice in which a mutant PolG (polymerase γ) introduces mutations in mtDNA at an increased rate so that animals expressing this protein accumulate mitochondrial mutations at a high frequency. These animals appear to age prematurely, developing age-related diseases early, losing their hair and greying prematurely [66,67]. This feat of genetic engineering has been met with a mixture of enthusiasm and scepticism. For example both Miller [68] and Khrapko et al. [69] have pointed out that the rate of mutations in these animals exceeds anything seen in aging in real life by orders of magnitude, and that there are other associated pathologies that appear in these animals that might make them sick, with very low red cell counts, intestinal damage etc. It is always hard to be sure when an animal ages or dies prematurely how good a model this is to relate to normal aging as it is always likely that non-specific pathological changes may shorten life.

Extending life span is perhaps more challenging, and this was achieved in the knockout mouse for the adaptor protein p66shc [70]. These animals showed a 30% increase in lifespan and their cells showed a resistance to oxidative stress. Remarkably, the implication is that we all express a protein which normally reduces our life span and drives aging. It also encourages the idea that aging is associated with accumulating oxidative damage.

Mitochondrial damage by changes in cell signalling and local environment

Mitochondria are subject to changes in their local environment, through changes in intracellular calcium signalling, changes in free radicals generated either by mitochondria themselves or by other mechanisms [XO (xanthine oxidase) and NADPH oxidase], and they are exposed to and are sensitive to changes in NO (nitric oxide) as it changes with cell signalling events. These are all part of the normal life of a mitochondrion inside the cell and, under normal physiological circumstances, may act as important signalling mechanisms driving altered mitochondrial function in response to altered cell physiology. The best characterized mechanism here involves changes in intracellular calcium. Once controversial, it is now clear that mitochondria will accumulate Ca2+ in response to a rise in cytosolic Ca2+ during routine cell signalling events: the Ca2+ signals that drive contraction of muscle (be it heart, skeletal or smooth) and the Ca2+ signals involved in secretion or those involved in neuronal signalling all cause a rise in intramitochondrial Ca2+ mediated by the mitochondrial calcium uniporter. Calcium is then returned to the cytosol by a sodium/calcium exchanger (mtNCX). It is now a widely held view that these changes in matrix calcium concentration are largely responsible for the up-regulation of OXPHOS that increases mitochondrial ATP generation in response to the increased demand that will inevitably accompany these physiological events. Ca2+ in the matrix activates rate-limiting enzymes of the tricarboxylic acid cycle, seems to increase ATP synthase activity (although the mechanism remains unclear [71]), and increased cytosolic Ca2+ also increases mitochondrial transporter activity [72,73]. NO will compete with oxygen at complex IV and so may inhibit mitochondrial respiration, especially at the low ambient pO2 that is probably present deep within most tissues [74]. The functional roles of this mechanism during normal NO signalling remains controversial, as does the specific role of intramitochondrial NO generation [75,76]. There is also some evidence that ROS generated by the electron transport chain may be important in some physiological signalling pathways (e.g. see [77]).

The mitochondrial permeability transition and disease

All of these mechanisms may become ‘unstuck’ during pathological conditions, where the combination of an excessive Ca2+ load, excessive free radical generation sufficient to induce a state of oxidative stress (a much abused term, which specifically implies an excessive oxidation that exceeds the antioxidant capacity of the cell or tissue), or excessive or inappropriate NO generation. The mitochondria under these circumstances seem to act as coincidence detectors, so that a combination of Ca2+ overload and oxidative stress will trigger mitochondrial dysfunction which may be fatal. The mitochondrial mechanism driving cell death under these circumstances is ascribed to opening of the mPTP (see Figure 1). The molecular identity of this pore has been a source of great confusion. Until recently, we thought we knew that this consisted of the ANT (adenine nucleotide translocase; in the inner membrane), VDAC (voltage-dependent anion channel; in the outer membrane) and CypD (cyclophilin D; in the matrix) which binds CsA (cyclosporin A) [7880]. The roles of ANT and VDAC have been thrown into question by the failure of knockout experiments to prevent mPTP opening [81,82], whereas the role of CypD has been convincingly confirmed using the same approach [8385]. It seems likely that the ANT is involved somehow in the modulation of pore opening rather than forming part of the pore, although it was established many years ago that the ANT can change conformational state to act as a pore in the presence of raised Ca2+ [86], repeatedly shown also for other inner membrane transporters [87].

At present then, the molecular identity of the pore remains uncertain. Spanning at least the IMM, pore opening causes a complete collapse of mitochondrial membrane potential (Δψm) and the loss of some matrix components, certainly a loss of Ca2+ and, indeed, it seems likely that NADH will also leak from the matrix, silencing respiration. The pore has such a large conductance that neither an increase in respiratory rate nor reversal of the ATPase can counter this leak. It seems that the pore can show brief transient openings which are harmless for the cell [88], but once pore opening is sustained, mitochondria cannot synthesize ATP and will even consume ATP as the ATPase runs ‘backwards’, and so cell death seems inevitable mostly through collapse of ATP. In cells that have a high glycolytic capacity and which can therefore maintain cellular ATP reserves, mitochondrial swelling and the translocation of pro-apoptotic factors may initiate the apoptotic cascade, although apoptosis induced by classical inducers (such as staurosporine) appears to be primarily independent of CypD and mPTP opening [83].

mPTP opening is inhibited by ATP and promoted by Pi and is most likely to occur under conditions of ATP depletion coupled with Ca2+ overload. These are precisely the conditions that prevail at a time of reperfusion following a period of ischaemia in the heart; at reperfusion, Δψm is restored at a time when ATP is depleted and Pi and Ca2+ are both high. What is perhaps most exciting about the pore is its clear involvement as a decisive event which defines the point of no return in the progression to necrotic cell death and also its modulation by pharmacological agents, classically CsA, while some newer agents are now beginning to appear as mPTP suppressors (see e.g. [89,90]). This raises the potential for therapeutic intervention for major diseases and represents perhaps the greatest translational triumph so far for mitochondrial-related disease. Predicted by Crompton and co-workers [91], protection of the heart by CsA specifically at the time of reperfusion was later demonstrated by Halestrap and others and now this approach is being applied in clinical trials (for a review see [92]). The same scenario of raised intramitochondrial Ca2+ and oxidative stress seems to be repeatedly recapitulated as a pathogenic mechanism. In the CNS a role for the mPTP has also been proposed in ischaemia and in the related glutamate excitotoxicity [93]. Ischaemic injury in the CNS is reduced in the CypD-knockout mouse [83], but evidence for mPTP opening to glutamate toxicity is not so strong and neurons cultured from CypD-knockout mice are not significantly protected from glutamate-induced cell death [94]. In the pancreas there is increasing evidence for a role of the mPTP in acute pancreatitis [95]. A most surprising role for the mPTP has been proposed for two human diseases involving a deficiency of collagen VI (known as Ullrich’s Congenital Muscular Dystrophy and Bethlem myopathy) [96]. It is not clear in any way how or why a mutation in an extracellular matrix protein should alter myocyte mitochondrial function, and the findings remain somewhat controversial [97,98]. Genetic ablation of CypD or pharmacological suppression of the mPTP seems to be effective in preventing muscle loss in animal models of the disease and clinical trials are in process in children with this horrible disease [89,99,100].

Role of the mitochondrial ATPase as an ATP consumer: pathological consequences

One issue that has specifically interested us in recent years has been the potential role of mitochondria as drivers of ATP consumption. This paradoxical role is based on the fundamental properties of the ATP synthase. The enzyme is in effect a membrane-bound proton-translocating ATPase which normally runs as an ATP synthase driven by the proton gradient established by respiration. However, it will operate in either direction, depending on the balance of free energies available from the phosphorylation potential and from the protonmotive force. Normally it will run in the synthetic mode, as the intramitochondrial phosphorylation potential is kept low (by the export of ATP via ANT) and the protonmotive force is high. If mitochondrial respiration is compromised, most simply by anoxia/ischaemia, but also by mutations, by secondary oxidant damage, by inhibition of cytochrome c oxidase by NO etc., and the protonmotive driving force fails, the enzyme will operate in reverse, consuming glycolytic ATP and pumping protons. Inhibition of respiration is accompanied by up-regulation of glycolysis, and the proton pumping may sustain or even increase the potential [101,102]. We have seen this mechanism in operation repeatedly in pathological models, in ischaemia in rat cardiomyocytes, Δψm is maintained by the ATPase which is a sufficiently powerful mechanism to deplete cellular ATP and precipitate cell death [103105]. The reversal of the ATPase and a role in maintaining Δψm is apparent in neurons with knockout of PINK1 (a protein associated with PD; see above), and in transmitochondrial cybrid cells carrying a mutation associated with a MELAS (mitochondrial encephalomyopathy, lactic acidosis and stroke-like episodes)-like syndrome [10]. While Δψm is maintained by this mechanism rather than by respiration, especially in the presence of oxygen, it seems likely that ROS generation will be increased, and as the sustained potential will still promote Ca2+ influx it is more likely that mitochondria may become Ca2+ overloaded in the face of pathology and cellular deregulation of Ca2+ homoeostasis. What is most interesting here is that a compound BMS-199264, which selectively inhibits ATPase without inhibiting synthase activity, reduces infarct size in Langendorff perfused rat hearts, suggesting potential therapeutic avenues for dealing with this mechanism [105]. Equally interesting is the fact that the ATPase activity is regulated by a protein, known as IF1 (the gene is ATPIF1), which inhibits this reverse activity. At present almost nothing is known about the relative expression level of this protein in different tissues, or about the regulation of expression, but this again will interfere with the ATPase activity and its contribution to pathophysiology needs further exploration (see [106] for a recent review).

Mitochondria in tumour formation and proliferation

Studies on altered cellular metabolism and mitochondrial function in tumour formation and cancer cell proliferation have recently gained an unexpected momentum, following the recognition of mitochondrial involvement in the development of at least two fundamental hallmarks of cancer: limitless proliferation and evasion of apoptosis [107,108]. Given the role of mitochondria in adapting cellular energetics in response to internal (genetically determined) and external (nutrient supply, metabolic stress) demands, it is perhaps evident that the organelle will also dictate differences between the metabolism of resting and proliferating cells. Glucose uptake is enhanced in many cancers (exploited by diagnostic imaging techniques; [109]), which primarily use glycolysis for energy production in contrast with the greater dependence of their non-transformed tissue counterparts on OXPHOS, even under conditions in which the oxygen supply is adequate to support the more efficient production of ATP by the mitochondria (Warburg effect; [110]). There is now substantial evidence indicating that enhanced glycolysis coupled with increased NADPH formation through the pentose phosphate shunt, and enhanced exit of citrate from the tricarboxylic acid cycle, together produce sufficient amounts of ATP and supply the necessary carbon source for lipid and protein biosynthesis for proliferating cells (for reviews see [111,112]). In addition, enhanced lactate production and extrusion which acidify the extracellular space promote tumour growth and invasion [113].

In his pioneering work Warburg also proposed that changes in the intermediary metabolism of proliferating tumour cells are triggered by primary changes in mitochondrial function, and this is still one of the much debated, and recently intensely studied, questions in the field of cancer metabolism [114,115]. Two lines of evidence support this view. First, both oncogenic and tumour suppressor pathways directly impinge on pyruvate entry into the tricarboxylic acid cycle and OXPHOS, either reducing or enhancing mitochondrial ATP production respectively [116119]. Moreover, mutations of the mitochondrial genome are frequently observed in tumour tissues [120] and were shown to drive metastasis formation [121]. These changes appear to promote cell proliferation either by promoting the Warburg effect or by enhancing cellular ROS production. However, it should be noted that (i) reduced OXPHOS may be secondary to hypoxia in the tumour environment, mediated by an evolutionarily conserved adaptive pathway, principally implemented by the oxygen-sensing HIF-1 (hypoxia-induced factor 1) [122,123], and (ii) mitochondrial cellular transformation can also be accompanied even by increased OXPHOS (see e.g. [124]). Thus further studies will be necessary to dissect primary and secondary mitochondria-related metabolic pathways in tumour cells with the promise of developing radically new approaches in cancer therapy [125,126].

The relationship between cell death and mitochondrial function is a similarly complex issue that we are just beginning to understand. Mitochondrial dysfunction is usually associated with cell death; for example, when accompanied by cellular Ca2+ deregulation in situations such as glutamate-induced neurotoxicity (see above). Under these conditions, mitochondrial biogenesis can play a protective role, e.g. induced by PGC-1α (peroxisome-proliferator-activated receptor γ co-activator 1α) in neurodegenerative disease [127129]. However, in cancer cells, the situation seems quite different and, instead, a reciprocal relationship between mitochondrial function and apoptosis has been found in several instances (for a review see [130]). Thus cancer cells, in which OXPHOS is impaired (see above), are more resistant to apoptosis induced either by stress (hypoxia, oxidative stress or nutrient depletion) or by chemotherapeutic agents (see e.g. [131]). This phenomenon could be explained by (i) the double-edged function of certain pro-apoptotic proteins [e.g. cytochrome c, and AIF (apoptosis-inducing factor)], which are required both for maintaining OXPHOS and triggering apoptosis, following their release from mitochondria [132]; or by (ii) the ATP-dependence of the apoptotic machinery [133]. Although none of these mechanisms has been compellingly demonstrated, the further investigation of these pathways holds an enormous potential for the development of alternative strategies in cancer therapy [134].

Mitochondria in diabetes

We have recently published a detailed account of the fundamental role of mitochondria in β-cell death in both Type 1 and Type 2 diabetes mellitus (T1DM and T2DM respectively; [135]). Intriguingly, similarly to cancer, mitochondria appear to represent a pathologically important hub of metabolic deregulation and cell death pathways also in diabetes. Indeed, combined mitochondrial apoptotic and necrotic cell death pathways, induced by cytokines, virus recognition and cellular stress pathways, are responsible for β-cell death in T1DM [136,137]. Furthermore, while mitochondrial dysfunction triggers a similar apoptotic/necrotic β-cell death in T2DM accompanying overnutrition-induced stress [138], it is also responsible for the functional derangement of GSIS (glucose-stimulated insulin secretion), reflecting the central role of mitochondria between glucose provision and insulin secretion [139,140].

Caspase-dependent β-cell apoptosis in T1DM is mediated by the combination of either IL-1β (interleukin-1β) or TNFα with IFN-γ (interferon-γ) [141], but can also be observed due to the induction of cellular stress responses, reflecting altered ER (endoplasmic reticulum) Ca2+ homoeostasis [142] and increased NO synthesis [143], examples of secondary (acquired) mitochondrial dysfunction. This can lead to depletion of the cellular ATP pool, accumulation of AMP, and cell death induction by overactivation of the AMPK (AMP-activated kinase pathway) [144]. Alternatively, apoptosis/necrosis can be induced by the depletion of the cellular NAD+/NADH pool mediated by PARP [poly(ADP-ribose) polymerase] activation and accompanied by AIF translocation from the mitochondrial intermembrane space to the nucleus [145,146]. Similarly, ER stress and NO signalling have also been found responsible for mitochondria-mediated β-cell death in T2DM, but evidence is accumulating that even primary mitochondrial dysfunction can link altered metabolism to the development of at least the Type 2 variety of diabetes [147]. Although the exact nature of the primary mitochondrial alterations has not yet been clarified, mechanisms similar to those that operate in increased OXPHOS-dependent cell death of cancer cells might be relevant in this case. Chronic exposure of pancreatic islets to elevated levels of nutrients induces β-cell dysfunction and ultimately triggers β-cell death, usually requiring both increased glucose and lipid exposure resulting in increased pyruvate supply to the mitochondria [148]. Moreover, this effect appears to be combined with an imbalance between the anaplerotic (pyruvate carboxylase-mediated) and oxidative (pyruvate dehydrogenase-mediated) metabolic routes [149]. Again, how these processes are linked to increased apoptotic sensitivity of the β-cell is poorly understood, but further studies on these fundamental mechanisms will most likely reveal alternative therapeutic targets of the disease.


Mitochondria are fundamental to cell life and to maintain normal physiological function. It is therefore hardly surprising that defects in mitochondrial function that are not lethal will cause disease, and this chapter indicates perhaps the breadth of that involvement and the array of major diseases in which mitochondrial dysfunction is implicated. As we begin to unravel the mechanisms underlying mitochondrial dysfunction in disease and new potential therapeutic targets are revealed, so the prospect of developing new therapeutic strategies becomes closer to a reality.


  • As mitochondria are fundamental to cell life, maintain normal physiological function and control pathways to cell death, defects in mitochondrial function cause cellular dysfunction or death underlying a wide range of diseases.

  • Mitochondria-related diseases can be triggered by inherited (primary) and acquired (secondary) defects in organelle function.

  • Mutations of both nuclear and mitochondrial DNA encoded proteins are the cause of primary mitochondrial diseases, following Mendelian or maternal inheritance patterns respectively.

  • Secondary mitochondrial diseases may result from impaired mitochondrial quality control. These include disorders of intramitochondrial protein folding, degradation, mitochondrial fusion and fission, and selective mitochondrial autophagy.

  • Both primary and secondary mitochondrial dysfunction manifest as altered intermediary metabolism of carbohydrates, lipids and proteins, redox imbalance and the tendency to undergo mitochondrion-mediated cell death (outer membrane permeabilization/apoptosis or permeability transition/necrosis). Together these lead to a failure of the cell to cope with environmental and intrinsic stressors.

  • Mitochondrial dysfunction underlies a wide variety of diseases including muscle and neurodegenerative disease, cardiovascular diseases, diabetes and cancer, providing an important potential therapeutic target for many current major diseases.


G.S. is supported by the Parkinson’s Disease Society (Project Grant number G-0905), and work in the laboratory of M.R.D. is supported by grants from the Wellcome Trust and Medical Research Council.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
  40. 40.
  41. 41.
  42. 42.
  43. 43.
  44. 44.
  45. 45.
  46. 46.
  47. 47.
  48. 48.
  49. 49.
  50. 50.
  51. 51.
  52. 52.
  53. 53.
  54. 54.
  55. 55.
  56. 56.
  57. 57.
  58. 58.
  59. 59.
  60. 60.
  61. 61.
  62. 62.
  63. 62a.
  64. 63.
  65. 64.
  66. 65.
  67. 66.
  68. 67.
  69. 68.
  70. 69.
  71. 70.
  72. 71.
  73. 72.
  74. 73.
  75. 74.
  76. 75.
  77. 76.
  78. 77.
  79. 78.
  80. 79.
  81. 80.
  82. 81.
  83. 82.
  84. 83.
  85. 84.
  86. 85.
  87. 86.
  88. 87.
  89. 88.
  90. 89.
  91. 90.
  92. 91.
  93. 92.
  94. 93.
  95. 94.
  96. 95.
  97. 96.
  98. 97.
  99. 98.
  100. 99.
  101. 100.
  102. 101.
  103. 102.
  104. 103.
  105. 104.
  106. 105.
  107. 106.
  108. 107.
  109. 108.
  110. 109.
  111. 110.
  112. 111.
  113. 112.
  114. 113.
  115. 114.
  116. 115.
  117. 116.
  118. 117.
  119. 118.
  120. 119.
  121. 120.
  122. 121.
  123. 122.
  124. 123.
  125. 124.
  126. 125.
  127. 126.
  128. 127.
  129. 128.
  130. 129.
  131. 130.
  132. 131.
  133. 132.
  134. 133.
  135. 134.
  136. 135.
  137. 136.
  138. 137.
  139. 138.
  140. 139.
  141. 140.
  142. 141.
  143. 142.
  144. 143.
  145. 144.
  146. 145.
  147. 146.
  148. 147.
  149. 148.
  150. 149.
View Abstract