Mitochondria in health and disease: perspectives on a new mitochondrial biology

Molecular Aspects of Medicine 25 (2004) 365–451 www.elsevier.com/locate/mam

Review

Mitochondria in health and disease: perspectives on a new mitochondrial biology Michael R. Duchen

*

Department of Physiology and Mitochondrial Biology Group, University College London, Gower Street, London WC1E 6BT, UK

Abstract The integrity of mitochondrial function is fundamental to cell life. It follows that disturbances of mitochondrial function will lead to disruption of cell function, expressed as disease or even death. In this review, I consider recent developments in our knowledge of basic aspects of mitochondrial biology as an essential step in developing our understanding of the contributions of mitochondria to disease. The identification of novel mechanisms that govern mitochondrial biogenesis and replication, and the delicately poised signalling pathways that coordinate the mitochondrial and nuclear genomes are discussed. As fluorescence imaging has made the study of mitochondrial function within cells accessible, the application of that technology to the exploration of mitochondrial bioenergetics is reviewed. Mitochondrial calcium uptake plays a major role in influencing cell signalling and in the regulation of mitochondrial function, while excessive mitochondrial calcium accumulation has been extensively implicated in disease. Mitochondria are major producers of free radical species, possibly also of nitric oxide, and are also major targets of oxidative damage. Mechanisms of mitochondrial radical generation, targets of oxidative injury and the potential role of uncoupling proteins as regulators of radical generation are discussed. The role of mitochondria in apoptotic and necrotic cell death is seminal and is briefly reviewed. This background leads to a discussion of ways in which these processes combine to cause illness in the neurodegenerative diseases and in cardiac reperfusion injury. The demands of mitochondria and their complex integration into cell biology extends far beyond the provision of ATP, prompting a radical change in our perception of mitochondria and placing these organelles centre stage in many aspects of cell biology and medicine.
2004 Elsevier Ltd. All rights reserved.

*

Tel.: +44-207-679-3207; fax: +44-207-813-0530. E-mail address: [email protected] (M.R. Duchen).

0098-2997/$ - see front matter
2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.mam.2004.03.001

366

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

Contents 1.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 367

2.

Evolutionary history and some basic biology of the mitochondrion . . . . . . . . . 2.1. Principles of mitochondrial bioenergetics . . . . . . . . . . . . . . . . . . . . . . . . 2.1.1. Structure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1.2. The chemiosmotic principle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2. Mitochondrial biogenesis and replication . . . . . . . . . . . . . . . . . . . . . . . . 2.2.1. The beginning: mitochondria at fertilisation . . . . . . . . . . . . . . . . . . . 2.2.2. PGC1a and the coordination of mitochondrial biogenesis . . . . . . . .

368 369 369 372 374 374 378

3.

Measurement of mitochondrial function in cells . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1. Measurement of mitochondrial membrane potential . . . . . . . . . . . . . . . . 3.1.1. Measurement of
wm (i) dye distribution and redistribution . . . . . . 3.1.2. Measurement of
wm (ii) the quench/dequench mode . . . . . . . . . . . 3.2. Assessing redox state in situ––measurements of NAD(P)H and flavoprotein autofluorescence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3. Location of proteins to mitochondria––using fluorescent mitochondrially targetted proteins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

379 380 380 383

Mitochondrial function and cell signalling . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1. Mitochondrial calcium uptake . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2. Mitochondrial Ca2+ efflux . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3. Mitochondrial influence on resting cytosolic calcium concentration–– the ‘set point’ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4. Mitochondria and [Ca2+]c microdomains . . . . . . . . . . . . . . . . . . . . . . . . 4.5. Mitochondrial Ca2+ uptake and mitochondrial function . . . . . . . . . . . . .

389 389 393

4.

384 387

393 394 397

5.

Mitochondria and free radical generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 398

6.

Uncoupling proteins and mitochondrial oxidative stress. . . . . . . . . . . . . . . . . . 400

7.

Mitochondria as sources and targets of NO . . . . . . . . . . . . . . . . . . . . . . . . . . 401

8.

Mitochondria and oxygen sensing, tissue oxygen tensions . . . . . . . . . . . . . . . . 403

9.

Mitochondria and glucose sensing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 406

10.

Mitochondria and cell death––necrosis, apoptosis and necroptosis? . . . . . . . . . 10.1. Mitochondria as ATP consumers . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.2. The mitochondrial permeability transition pore (mPTP) . . . . . . . . . . . . 10.3. Apoptosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

407 408 411 415

11.

Mitochondria and neurodegenerative disease . . . . . . . . . . . . . . . . . . . . . . . . . 11.1. Mitochondrial calcium overload as trigger of neuronal death . . . . . . . . . 11.2. Inducers of neuronal damage: the role of free radical species . . . . . . . . . 11.3. Motor neuron disease. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.4. Alzheimer’s disease and the toxicity of amyloid beta protein . . . . . . . . .

418 419 421 422 424

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

Ab and [Ca2+]c signalling. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mitochondria and oxidative stress as mediators of amyloid toxicity . . . . . . . . . . . . . . . . . . . . . . . . . . . . Huntington’s disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Parkinson’s disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

11.4.1. 11.4.2. 11.5. 11.6.

367

425 427 432 432

12.

In the heart: mitochondrial calcium overload and reperfusion injury. . . . . . . . . 433 12.1. The ‘mitochondrial KATP channel’: a target for cytoprotection . . . . . . . . 435

13.

Postscript . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 437

Acknowledgements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 438 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 438

1. Introduction The mitochondrion lies at the heart of cell life and cell death. To put these structures in context, consider that every child knows that we must breathe oxygen to stay alive. And why? Because our mitochondria demand oxygen in order to function. 98% or so of the oxygen that we breathe is destined to be consumed by our mitochondria. Without mitochondria, oxygen would be of no use to us, we would not need the oxygen transfer machinery of the lungs. There would be no need for red cells, or haemoglobin, or even for the complexity of a circulatory system to deliver oxygen to the tissues. Even much of the machinery required to acquire and deliver carbon substrate to the tissues is delivering carbon to the mitochondria for oxidative phosphorylation. Consider then how much of the physiology of higher organisms is dictated by the demand of our mitochondria for a supply of oxygen. And yet without the supply of energy provided by mitochondria, the evolution of higher organisms would not have been possible. Mitochondria are essential to maintain the battle against entropy that is necessary to sustain life. They provide the energy required for almost all cellular processes––to allow muscle to contract––and that includes skeletal muscle, cardiac muscle, the smooth muscle of the gut, of the vasculature, of the lungs. They are required to provide energy to maintain ionic gradients across cell membranes, necessary for the excitability of excitable cells, to allow the accumulation of secreted material into vesicles, and to permit the vesicle fusion and cycling necessary for the secretion of hormones, of neurotransmitters––all the processes central to sophisticated life. And so it follows inevitably, that mitochondrial dysfunction will lead to disease, ranging from the subtle alterations in function in tissues that may manifest as disease and illness, to major defects in tissue function that may lead to major handicap or death. In fact, mitochondria are essential not just as providers of energy, but they also are exquisitely and intimately involved in a subtle discourse with other aspects of cell physiology, with cellular calcium signalling in

368

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

particular, so fundamental an aspect of physiology that again, any defect is bound to lead to dysfunction and illness. Mitochondria are damaged during the reperfusion of ischaemic heart, brain and kidney; mutations of mitochondrial proteins gives rise to a range of ill understood patterns of disease; mitochondrial dysfunction has been implicated in all the major neurodegenerative diseases––Parkinson’s, Alzheimer’s, motoneuron-disease (LouGehrig’s disease or Amyotrophic Lateral sclerosis) and possibly in Multiple Sclerosis. Mitochondrial dysfunction may give rise to cardiomyopathy. Mitochondrial function is central to transduction in the pancreatic beta cell and the secretion of insulin in response to glucose, and so dysfunction gives rise to diabetes. Most recently evidence points to a major role for mitochondrial dysfunction in the condition of multi-organ system failure in sepsis. It is a long, and growing, list. Accumulations in mitochondrial defects have been implicated as a mechanism of ageing and age related disease. Indeed, the production of free radicals by mitochondria has been considered by many a key in the cellular injury that appears to underlie the process of ageing, while some of the genes identified in the control of longevity appear to target mitochondria or at least to alter antioxidant defences of the cell. Over the last few years, the central role of mitochondria in the regulation of organised or programmed ‘apoptotic’ cell death has been a revelation. The centrality of apoptosis in development and in disease is extraordinary. If the process of apoptosis is activated inappropriately it will lead to tissue dysfunction and damage, and if it fails to be activated when necessary, cancers are the result, making this whole area absolutely critical in understanding the role of mitochondria in disease. And so mitochondria and mitochondrial dysfunction have been implicated in many different aspects of health and disease. I believe firmly that understanding of complex interactions between mitochondria and other aspects of cell function in pathophysiological states requires as full a general understanding of all aspects of mitochondrial biology as possible. In this review, I will therefore consider some of these fundamental attributes of mitochondria and consider the role of mitochondrial dysfunction in generating disease. I will particularly try to draw attention to the some of the more subtle aspects of mitochondrial biology that we are only just beginning to understand, that may well also impact on tissue function and on wellbeing.

2. Evolutionary history and some basic biology of the mitochondrion The mitochondrion is thought to have evolved from a bacterial progenitor around the same time and in much the same way as the other major residual symbiotic cellular structure, the chloroplast. It is probably no exaggeration to suggest that the whole evolution of sophisticated life forms has been dependent on this evolutionary step, the chloroplast harnessing energy from the sun to generate oxidiseable fuels, and the mitochondrion to use those fuels to generate energy rich biochemical compounds.

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

369

The primary remaining evidence of their independent bacterial origin is the presence of mitochondrial DNA (mtDNA), a simple circular genome that has structural characteristics very similar to the circular DNA of primitive bacteria (Leblanc et al., 1997). Mitochondria and chloroplasts are the only structures in cells to possess their own DNA which is distinct from the somatic DNA of the nucleus. However, mitochondrial DNA encodes only a fraction of the proteins that are fundamental for mitochondrial function, while the vast majority of proteins required to build mammalian mitochondria are encoded in the nucleus and transported to the mitochondria. Thus, human mtDNA is a circle of only 16.6kb double stranded DNA encoding just 13 proteins of which all are components of the respiratory chain. In addition, some 24 other genes are present on the mtDNA, encoding two rRNAS and 22 tRNAs required in turn to synthesise the 13 proteins. It is not feasible to construct so complex an organelle from so few proteins, and a total of about 850 proteins are in fact required, the majority encoded in the nuclear DNA (nDNA). These structural considerations immediately pose fascinating questions about the coordination of mitochondrial growth and replication, the matching of gene products of mitochondrial and nuclear origin, the trafficking, targeting and import of proteins to mitochondria, some of which will be considered later. When a system is poised so precariously on the dictates of such a complex coordination, does it not seem inevitable that there will be examples in which the system fails and results in pathology? I will discuss the issue of mitochondrial replication and biogenesis and the consequences of mitochondrial mutations below, as I consider it necessary to first understand the principles that govern mitochondrial function in order to understand the functional impact of mutations and protein expression. 2.1. Principles of mitochondrial bioenergetics 2.1.1. Structure In order to make sense of the literature dealing with mitochondrial function, it seems to me necessary to have a reasonable grasp of the basic principles of mitochondrial bioenergetics. These have been reviewed many times by many authors in recent years (Duchen, 1999; and see Ferguson and Nicholls, 2002) but I will give a brief overview here. Mitochondria are bounded by two membrane systems, the inner and outer mitochondrial membranes which occasionally come together, rather like gap junctions, to form junctional complexes or contact sites. The inner mitochondrial membrane is largely impermeant and forms the major barrier between the cytosol and the mitochondrial matrix. The space between the two membranes is referred to as the intermembrane space, itself recently revealed as an interesting microenvironment of which we understand little, but housing some proteins which play major roles in cell physiology, in mitochondrial energetics and in cell death––most notably perhaps cytochrome c and creatine kinase. The properties of the outer membrane are less well understood. It has often been assumed to be relatively permeant to small molecules and ionic species, but recent studies suggest that it might form a more important and potentially regulated barrier––see below.

370

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

The inner mitochondrial membrane is usually depicted forming multiple infoldings which form the cristae which house a variety of mitochondrial enzymes systems which are membrane bound. The structure of cristae seems to vary enormously between different tissues, and the functional significance of these differences in structure remains largely mysterious. Thus, in EM fixed specimens from brown fat (see section on uncoupling proteins) mitochondria appear tightly packed with lamellar cristae, while steroid synthesising tissues have cristae that seem to form tubular networks. The latter is particularly intriguing as the rate limiting enzymes of steroid synthesis are housed in mitochondria (one wonders why?). Recently, several groups have attempted to produce detailed three-dimensional reconstructions of cristae structure (Fig. 2A) and to model the bioenergetic consequences of changing shapes of cristae (Frank et al., 2002; Hsieh et al., 2002; Mannella et al., 1997; Mannella et al., 1998; Deng et al., 1999; Frey and Mannella, 2000). Recent detailed EM reconstructions suggest that mitochondrial cristae do not in fact form the neat parallel folds so often shown in the textbooks, but rather form complex arrays of tubules, which may even be continuously fusing and dividing, and which may not even always show continuity with the intermembrane space (Renken et al., 2002; Perkins et al., 1997, 1998; Perkins and Frey, 2000; Perkins et al., 2001; Duguez et al., 2002; Griparic and Van Der Bliek, 2001). It has to be said that we still have no real understanding of the functional significance of these features and variations of detail of mitochondrial structure. More ’macroscopic’ features of mitochondrial structures also vary between tissues. Mitochondria clearly form complex highly mobile interconnecting networks in some tissues although they seem to exist as more structurally discrete and independent structures in other tissues. Once again, the functional consequences of these differences have yet to be explored adequately. It has been suggested that mitochondrial networks may form a compartment in which contents and membranes are constantly interchanging, so that perhaps it would be misleading to talk of ‘a mitochondrion’. Some recent studies suggest that mitochondria may form an electrically contiguous network, in which soluble messengers, such as calcium, may be able to diffuse extensively (e.g. see Rutter and Rizzuto, 2000). Other studies fail to support this concept (Nitschke et al., 1997; Collins et al., 2002; Collins and Bootman, 2003). However, most such models of mitochondrial networks have been generated from cell lines maintained in culture. Even within different forms of primary cells in culture or freshly isolated cells, mitochondria look quite different between cell types; see Fig. 1. Perhaps these features vary between cell types, perhaps even between cells at different stages of the cell cycle. Perhaps those differences might even prove functionally important in defining the responses of different tissues to toxic processes. In fact, even within single cells there are suggestions of mitochondrial heterogeneity––in cardiac and possibly also in skeletal muscle, mitochondria appear to be composed of two pools, the interfibrillar mitochondria, which appear round and dense in electron micrographs compared to the paler, lighter mitochondria that lie underneath the plasma membrane, the subsarcolemmal mitochondria. These two classes of mitochondria require different procedures to isolate them and purified preparations of mitochondria contain two classes of organelle which have different

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

371

Fig. 1. Some comparative cell biology of mitochondrial structures. These images have been chosen to illustrate the variety in form and structure of mitochondria in a range of different cell types. In most of these images, mitochondria have been labelled with the potential sensitive indicator, tetramethyl rhodamine methyl ester (TMRM) except panel F in which mitochondria in a HeLa cell have been transfected with GFP. In A, note how the mitochondria in a neuron (in which the cell body is stained with calcein) are very small fine structures that may be difficult to resolve. In a lymphocyte (B) an astrocyte (C) and the HeLa cell (F) mitochondria form a filigree network of fine interconnected structures. In an adult freshly dissociated ventricular cardiomyocyte (D) the cell is densely packed with mitochondria arranged in rows along the length of the cell. Mitochondria are seen as fat bullet shaped structures seen in a neonatal cardiomyocyte in culture (E). The image in F was kindly provided by Mariusz Wieckowski.

biochemical properties, express different levels of enzymes, and appear to respond differently to stresses such as anoxia and reperfusion (see Lesnefsky and Hoppel, 2003). Again, the functional significance of this differentiation is really not understood, and the difference raises some fascinating issues about mitochondrial

372

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

replication, biogenesis and specialisation––what signals govern the differentiation of mitochondria into these separate pools? There appear to be differences in the content of respiratory enzymes––how are nuclear encoded proteins differentially directed to each population? It has been suggested that mitochondria in the immortalised HeLa cell line may show differences in potential, although this remains contentious (see below). HeLa cell mitochondria also are astonishingly mobile and are constantly fusing and separating, but such mobility is not seen in many primary cells in culture. This must prompt questions for which I believe we still have no answer: ‘What is the functional significance of differences in mitochondrial structure in different cells or even in mitochondria within the same cell? What are the consequences for cell function or for responses to pathophysiological conditions? To what extent is an experimental model developed in one cell type referable to another?’ 2.1.2. The chemiosmotic principle The key enzymatic components of the mitochondria are the citric acid or tricarboxylic acid (TCA) cycle and the respiratory or electron transport chain. The enzyme system of the TCA cycle breaks down carbon substrates acetyl CoA, derived from pyruvate, fatty acid and amino acid breakdown to generate CO2 and in the process to reduce NADþ to NADH and FAD2þ to FADH. These intermediates provide reducing equivalents to the respiratory chain which consists of a series of enzyme systems coupled together and described as Complex I, II, III and IV. Complex I, NADH dehydrogenase, Complex II succinate dehydrogenase, Complex III (ubiquinol cytochrome c reductase), and Complex IV (cytochrome c oxidase) are all complex membrane spanning enzyme systems consisting of many protein subunits. Interestingly, the proteins of Complex II are entirely encoded by the nucleus, while all the other complexes represent a mixture of proteins some encoded by nuclear DNA, some encoded by the mitochondrial DNA. Structural mechanistic features of each of the complexes have been resolved, but the roles and significance of some of the subsidiary components remain obscure––one can only presume that they must serve some modulatory function. It is most striking in the case of complex IV, in which there are 13 subunits, some of which vary between tissues (Anthony et al., 1990). See http://www.bmb.leeds.ac.uk/illingworth/oxphos/index.htm. Operationally, the key point is that there is a transfer of energy between the intermediates of the chain moving progressively energetically ’downhill’ from a reduced to an oxidised state. Electrons are transferred from NADH and FADH2 to Complexes I, and Complex II respectively (Fig. 2B and C), and these each then transfer electrons to ubisemiquinone which shuttles electrons to Complex III. Cytochrome c then shuttles electrons to Complex IV. Complexes I, II and III all contain non-haem iron, complex IV contains copper, while III, IV and cytochrome c all contain haem. All of these play a major role in catalysing the redox reactions. The reactions of I, III and IV are all also associated with the transfer of protons across the mitochondrial inner membrane, from matrix into the inter-membrane space, and this is key to establishing an electrochemical proton gradient, a ’proton-motive force’ predominantly expressed as a mitochondrial transmembrane potential, usually estimated at 150 to 180 mV negative to the cytosol. The mitochondrial membrane

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

373

Fig. 2. Structure and function of the mitochondrion. (A) (i) and (ii) show 3-D reconstructions of mitochondrial structures from EM tomography (with thanks to Guy Perkins, Terry Frey and Don Fox). The mitochondrion in (i) came from the dendritic tree of a cerebellar Purkinje cell. The spots in (ii) show contact sites between inner and outer membranes of rod photoreceptor mitochondria. In (B) is shown a cartoon to represent the functional components of the mitochondrion and (C) shows a more detailed scheme of the respiratory chain (see text for a fuller explanation).

potential lies at the heart of all the major bioenergetic functions of the mitochondrion, from the manufacture of ATP to accumulation of calcium, as it provides a force that drives the influx of protons or of calcium that simply move into the mitochondria down their electrochemical potential gradients. The influx of protons is predominantly directed through the proton channel of the F1 F0 -ATP synthase, a proton translocating ATPase that is driven ‘backwards’ by the proton gradient. The inward movement of protons (tending to depolarise the mitochondrial potential) drives a motor of the enzyme to phosphorylate ADP and to release ATP, beautifully demonstrated by Walker and his colleagues (Stock et al., 1999). ATP is then transported out of the mitochondria by the adenine nucleotide translocase (ANT) an electrogenic transporter that exchanges ADP for ATP. The inward transport of ATP

374

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

yields a net inward positive charge and so again tends to depolarise the mitochondrial potential slightly. The actions of the respiratory complexes as proton translocators also means that the respiratory rate is regulated by the proton gradients––respiration will run faster when the membrane is depolarised as a thermodynamic consequence of the reduced energy required to move protons, and will run more slowly if the membrane potential increases or hyperpolarises. This is readily demonstrated, as uncouplers, drugs (such as FCCP, CCCP and DNP) which shuttle protons across the membrane and allow the potential gradients to dissipate, promote an increase in respiratory rate, while inhibition of proton flux through the ATP synthase (using oligomycin, which effectively blocks the proton channel of the ATP synthase) increases the membrane potential and slows the respiratory rate. In this way, mitochondrial respiratory rate and the rate of ATP synthesis are coupled through a proton gradient even though they are in effect distinct and separate processes. The activity of cytochrome oxidase may also be regulated by the ATP/ADP ratio by an allosteric regulation at one of the may subunits on the protein complex (Arnold and Kadenbach, 1997; Follmann et al., 1998; Kadenbach et al., 2000). It is important to understand that mitochondrial membrane potential is normally maintained by cellular respiration. A loss of mitochondrial potential may be a reflection of several different mechanisms––an inhibition of respiration, a failure of provision of substrate or some kind of uncoupling mechanism that shunts the proton circuit and so dissipates the potential. The rate at which mitochondrial potential is dissipated in response to these different processes may also vary considerably between cells––we have found that in some cells, anoxia causes rapid collapse of potential, while in others, the potential is lost only slowly (see below, and Fig. 8). In part, this probably reflects differences in the native proton leak of different cell types, but in truth we as yet know little about the comparative cell biology involved. 2.2. Mitochondrial biogenesis and replication 2.2.1. The beginning: mitochondria at fertilisation Dogma says that all our mitochondria are maternally inherited. In principle, then, all mitochondria with all specialised features that we have mentioned are derived from prototypical mitochondria in the egg, which appear as simple structures far removed from the complex networks seen in many mature cells (Fig. 3A). In fact, while there is no doubt that this is generally true, there is some recent evidence for the penetration of the paternal mitochondrial genome in some patients who have inherited mitochondrial defects from the father (Schwartz and Vissing, 2002). It seems likely that this is an exception rather than the rule, and in principle, however, in most of us, all the mitochondria of the egg must differentiate into the various specialised structures found in all tissues. Indeed, there is some evidence that paternal mitochondria, arriving at the egg with the spermatozoon (Fig. 3B), are actively destroyed by the egg (Sutovsky et al., 2000). This involves the internalisation of the sperm mitochondria into the egg at fertilisation, followed by its ubiquitinisation and destruction. It is not entirely clear from an evolutionary perspective why

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

375

Fig. 3. Mitochondria at the beginnings. (A) a mouse oocyte shows many discrete round mitochondrial structures that seem to cluster in parts of the egg, in this case stained by a mitochondrial targeted GFP (mRNA was made and injected into the cell by Remi Dumollard). (B) In sperm, the mitochondria are closely packed around the tail, just behind the head of the sperm––here labelled with TMRM. (C) mitochondrial DNA is packed into ‘nucleoids’ which have been labelled here by a GFP tagged ‘TWINKLE’ (see text), (the image was kindly provided by Hans Spelbrink). The protein associates with the nucleoid and is essential for packaging the mtDNA, but also gives an invaluable way of following nucleoid structures experimentally. (D) illustrates the merging of mitochondria from HeLa cells which had been transfected with either GFP or DsRed and visualized (i) immediately or (ii) 8 h after cell fusion with PEG. Fused cells were maintained in cycloheximide to inhibit the synthesis of new nuclear encoded proteins. The double-labeling of polykaryon mitochondria demonstrates the exchange of fluorescent matrix proteins upon fusion of mitochondrial double membranes (kindly provided by Manuel Rojo). (E) shows the effect of drp1 on the structure of HeLa mitochondria. Control mitochondria are shown in (i) as a 1.5 lm slice after deconvolution of an image stack. The image shown in (ii) comes from cells transfected and therefore ovexpressing drp1, while those shown in (iii) were transfected with construct with amputation in the GTPase component of the drp1 (TRexDrp-1/k38a) that therefore acts as a dominant negative. (Image kindly provided by Gyorgy Szabadkai).

376

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

this should be necessary. One speculation perhaps is that spermatozoa have a very high turnover and therefore perhaps an increased probability of developing mtDNA mutations, compared to the egg which never divides following terminal differentiation, so that this destruction may be a mechanism to limit the occurrence of mitochondrial mutations. A counter argument is that mitochondria that stay within an egg for the lifetime of the woman may be exposed to a lifetime of radical species and so may be more prone to oxidative damage––indeed it has been suggested that this might be one basis for the reduced fertility seen in older women. Be that as it may, the essentially maternal inheritance of the mitochondria has obvious major implications for heritable mitochondrial diseases. Each cell may contain thousands of copies of mtDNA––the estimated numbers vary considerably (Reynier et al., 2001; Miller et al., 2003). A discussed above, it is not even clear if the term ‘a mitochondrion’ means anything, as mitochondria in many cell types may form a network that is constantly remodelled, although in other cell types the mitchondria may be more fixed as entities (e.g in muscle cells). The division and replication of mitochondria must in some way be linked to the division and replication of nuclear DNA, and is certainly under control of the nucleus (see below). A study by Davis and Clayton (1996) suggested that new mtDNA tends to be made close to the nucleus and then to radiate outwards through the cell, while in enucleated cells, mtDNA replication simply halted. In cells that show growth arrest (such as PC12 terminally differentiated into neurons) mitochondria close to the nucleus still show some incorporation of new nucleic acids which then appear later out in the cell periphery. The mitochondrial DNA appears to be packaged as clusters of mtDNA molecules within structures known as nucleoids. The structure of the nucleoid is well defined for yeast, but is less certain for mammalian mtDNA. Recent studies of a protein called TWINKLE have helped to identify the structure to some extent (see Fig. 3C). TWINKLE acts as a helicase, important in the maintenance of mtDNA integrity, and so binds to mtDNA. Labelled TWINKLE (Fig. 3C) can be seen localised to structures that seem consistent with nucleoids, small brightly staining spots within the mitochondrial network (Korhonen et al., 2003; Spelbrink et al., 2001). Immunofluorescence staining of TWINKLE or labelling with a GFP shows multiple tiny structures within individual mitochondria that colocalise with a mitochondrial transcription factor (Tfam; and see below) giving us tools that may allow further study of mitochondrial DNA replication in situ. What is also fascinating is that a mutation in TWINKLE has been linked to a human disease of progressive external ophthalmoplegia (Spelbrink et al., 2001), highlighting the importance of the maintenance of mtDNA in cell and tissue function, although why on earth such a mutation should specifically cause paralysis of the external ocular muscles seems a complete mystery. Currently available evidence suggests that the proteins found within mitochondria are largely shared throughout the mitochondrial population in the cell (although how is this consistent with the existence of two pools of mitochondria within single cardiocytes?). Thus, in a beautiful paper, Legros et al. (2002) have shown that fluorescent proteins targeted to mitochondria of one cell redistributed throughout the mitochondrial population of another cell when the two cells were fused. In these

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

377

experiments, fluorescent proteins targeted to mitochondria and fluorescing at different wavelengths were expressed in different cells and the cells were then fused. As the fluorescent proteins are mitochondrially targeted but expressed in the nucleus, the expression of new nuclear encoded protein was suppressed with cycloheximide. Over a period of time after fusion, the entire population of mitochondria gradually showed expression of both red and green fluorescent signals, appearing yellow, demonstrating the mixing of the proteins within the mitochondria (see Fig. 3D). The spreading and mixing of the fluorescent proteins through the mitochondrial population was dependent on the expression of the gene product Mfn 1 (mitofusin), and was prevented if Mfn1 expression was suppressed by antisense oligonucleotides, while expression of the dynamin related protein, drp 1 instead limited dispersion of the markers and caused mitochondrial fragmentation. These two proteins, probably representing more extensive families of proteins, therefore seem to regulate mitochondrial fusion and fission respectively, but their regulation in relation to the cell cycle is not yet understood. In Fig. 3E is shown recent experimental data (kindly provided by Gyorgy Szabadkai) to show the impact of these processes on mitochondrial structure in HeLa cells. The cells have been transfected with either drp1 (3Eii) or with a dominant negative mutant of drp1 which alters the GTPAse binding (3Eiii). The drp1 overexpression clearly causes mitochondrial fragmentation which the dominant negative is similar to overexpression of Mfn1, causing the appearance of long fibrillar mitochondria and clumping structures. Functional ‘complementation’––the sharing of mitochondrial gene products between mitochondrial structures––has also been demonstrated in a mouse generated with mutations of cytochrome oxidase in the mitochondrial pool (Inoue et al., 2000). However, the animal showed some heteroplasmy––a mixture of mutant and normal mtDNA––and the mutation contained a deletion so that it failed to encode a normal cyt c oxidase. Electron microscopic immunocytochemistry revealed that all mitochondria in the tissue appeared to express a normal cytochrome oxidase (Nakada et al., 2001), suggesting that the mtDNA, or, at least, the protein from normal mitochondria had been transferred to those with the deletion mutation. The regulation of biogenesis and replication of mitochondria remains a fascinating issue that must prove significant in defining the mitochondrial contribution to disease. The issue is complex and still poorly understood. Consider, though: cells of the GI tract have a rapid turnover and a high energy requirement and are packed with mitochondria. These cells replicate at a high rate, and so must their mitochondria, or they would become depleted of mitochondria. In contrast, in a cell like an oocyte or a neuron, which never replicates after reaching the terminally differentiated state, mitochondria too must remain at a steady state, either not replicating at all or at least maintaining a balance between degradation and replication. Even here, we do not know whether mitochondria simply remain in the cells throughout life or whether there is a continual turnover. The genetics of mitochondria follow population rather than Mendelian patterns. Thus, if there is a mutation of a single copy of the mitochondrial DNA in the egg, this may be transmitted into developing tissues, which will therefore be heteroplasmic–– they will have copies of mtDNA that are different in different populations of

378

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

mitochondria––and will have cells which contain both normal and mutant mtDNA. It is still not clear how mitochondria are segregated at cell division, how they distribute into growing tissues, or how their replication within those tissues is regulated. Thus, some of the mitochondrial myopathies are characterised by patchy distribution of muscle fibres which are packed with abnormal mitochondria, while in other fibres, the mitochondria are predominantly normal. It is often assumed to result from the segregation of one clonal selection of abnormal mitochondrial DNA in the diseased cell and somehow that abnormal mitochondrion has some selective advantage over the normal mitochondria such that it replicates faster and so becomes dominant. The consequence of this genetic behaviour is that patients with identical mutations of mtDNA may present with widely differing patterns of disease expression––there are examples of patients with identical mutations of Complex IV, some of whom present late in life with mild muscle weakness and others are debilitated early in life with severe stroke like episodes and lactic acidosis. This variability in clinical presentation is simply not understood, nor is the phenotypical expression of disease. It is open to speculation that perhaps interaction of disordered mitochondrial proteins with specific nuclear encoded mitochondrial proteins may be critical in defining the pattern of disease in different patients or in different tissues. Until we have good animal models to study, it seems almost impossible to address these issues adequately experimentally. 2.2.2. PGC1a and the coordination of mitochondrial biogenesis As indicated several times already, mitochondrial replication requires the complex coordination of mitochondrial DNA replication, coupled to the increased synthesis of proteins encoded by both the nuclear and the mitochondrial genomes. The balance of respiratory proteins must be maintained or function is compromised, and so the generation of both mitochondrially and nuclear encoded proteins must be synchronised. This appears to be achieved through the coordinated expression of a group of recently identified transcription factors. The key seems to be the transcription coactivator, PGC1a (peroxisome proliferator activated receptor c coactivator 1a). Much of the normal regulation of this pathway still needs to be elucidated. The most productive model used so far derives from the process of adaptive thermogenesis, in which adipocytes of brown adipose tissue in rodents increase mitochondrial mass and mitochondrial protein production in response to cold. This requires the increased synthesis of nuclear encoded mitochondrially targeted proteins, apparently achieved largely by PGC1a mediated upregulation of two transcription factors known as NRF-1 and 2 (nuclear respiratory factors 1 and 2). These also increase the transcription of another protein, mtf1 (mitochondrial transcription factor 1, also referred to as Tfam), a protein that transfers from the nucleus to the mitochondria where it promotes the increased production of mitochondrially encoded proteins and the replication of mitochondrial DNA. This system thus appears to coordinate the rate of synthesis of protein from the two genomes. What remains unclear is how this pathway operates in different cell types during growth and differentiation or in relation to the cell cycle, although we do know that PGC1a may be activated in brown adipose tissue by adrenaline acting via CREB activation. One fascinating example in which the upregulation of mitochondrial mass and function has been characterised is in muscle,

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

379

in which repeated stimulation causes an increase in mitochondrial mass, mtDNA copy number and mitochondrial protein expression and an increase in mitochondrial respiration. The process appears to be mediated via the sustained rise in intracellular calcium that accompanies repetitive stimulation of muscle which acts at CaMKinase IV, which in turns increases the activation of PGC1a. The import of nuclear encoded, mitochondrially targeted proteins is yet another important aspect of mitochondrial biology which is now becoming clearer. Mitochondrially targeted proteins express a presequence which is recognisable by a pair of channel proteins that span the outer and inner mitochondrial membranes––TOM (translocase of the outer membrane) and TIM (translocase of the inner membrane) respectively. These seem to act as a cog system to transfer proteins across the double membrane system. Precursor proteins are recognized by outer membrane receptors dependent on their internal targeting sequences in complex with cytosolic chaperones and dock to the receptor Tom70. After crossing the outer membrane through the TOM complex, the insertion into the inner membrane (a process dependent on the mitochondrial membrane potential) is mediated by the two translocases dependent on the type of precursor protein. Translocation of matrix proteins is completed by the action of the ATP-dependent heat shock protein, matrix Hsp70, sometimes called the import motor complex, accompanied by the removal of the targetting presequence. Hsp 60 is then involved in aiding the correct folding and localisation of the proteins, which is essential to ensure correct function (see Frazier et al., 2003; Voos, 2003; Pfanner and Truscott, 2002). Clearly, coordinated protein import is essential for mitochondrial maintenance and biogenesis, and it seems once again inevitable that disturbances in any of these complex systems or their coordination will generate disease. There is a nice illustration of mitochondrial division and fusion watched in a living HeLa cell on-line provided kindly by Mariusz Wieckowski (see http:// www.physiol.ucl.ac.uk/research/duchen_m/).

3. Measurement of mitochondrial function in cells In our attempts to understand aspects of mitochondrial function in the context of cellular physiology and pathophysiology, it is essential to have tools that permit the study of mitochondrial function within intact cells and that provide unambiguous data to inform us about the fundamental mechanisms involved in processes so central to cell and tissue function and survival. The analysis of the biochemical and bioenergetic principles that underpin the standard biochemistry textbook description of the mitochondrion as the ’powerhouse of the cell’ were all based on studies of mitochondria isolated in bulk from tissues such as the liver, the heart or the brain. Mitochondria were purified, and biochemical functions such as the rate of oxygen consumption, the rate of production of biochemical intermediates etc. were assessed. While such work provides the mainstay for our understanding of mitochondrial bioenergetics, it also has its limitations. Most obviously, under these conditions, it is impossible to ask questions about the specialisation of mitochondrial function in different cell types, where the yield of cells or of mitochondria might be limited. It is

380

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

certainly not possible to ask questions about spatial and temporal changes in mitochondrial function in single cells during events associated with cell signalling or in response to pathophysiological conditions. It is these kinds of questions that have driven the development of approaches which allow us to study mitochondrial function within individual cells. The available methodologies used to follow mitochondrial function have developed dramatically in recent years along with the growing perception of the central role of mitochondria in the life of the cell. However, many of these approaches are open to misinterpretation, confusing our attempts to unravel the key events in the progression to cell death (see Fink et al., 1998; Ward et al., 2000). The growth of fluorimetry and fluorescence imaging technology has enabled study of mitochondrial function within cells, either in populations or at the level of the single cell. In particular, this approach has allowed the study of mitochondrial function in individual cells in relation to other aspects of cell physiology or pathophysiology, and it has also made accessible questions about the comparative physiology of mitochondrial responses between different cell types. We have recently written at some length about the technology and possibility of measurement of mitochondrial function (Duchen et al., 2003), but will recapitulate some of these principles here as it seems important to understand, interpret and use these approaches intelligently. 3.1. Measurement of mitochondrial membrane potential Dwm has been assessed for many years by following the distribution of lipophilic cations, traditionally tetraphenyl phosphonium (TPPþ ). In a preparation of isolated mitochondria, TPPþ will accumulate into mitochondria on the addition of substrate. Following the generation of a potential, the concentration of the ion in the extramitochondrial space falls and this can be followed using an ion sensitive electrode manufactured with an appropriate semipermeable membrane. Fluorescence technology exploits the same principle, by following the (re)distribution of fluorescent lipophilic cations. Many fluorescent compounds, most notably rhodamine 123 (Rh123), tetramethylrhodamine ethyl and methyl esters (TMRE and TMRM, respectively), 5,50 ,6,60 -tetrachloro-1,10 ,3,30 -tetraethylbenzamidazolocarbocyanine (JC1), DiOC6 and DASPMI are both cationic and membrane permeant. They cross the cell membrane easily and partition between cellular compartments in response to the standing electrochemical potential gradients. Considerable confusion has arisen in the literature because these dyes have been used in cells different ways by different groups. It is essential to understand the significance, application and limitations of each approach if experimental design is to be accurately interpreted. Essentially, I shall refer to these approaches as the ‘redistribution method’, and the ‘quench/dequench method’. 3.1.1. Measurement of Dwm (i) dye distribution and redistribution For the rigorous measurement of mitochondrial potential, any of the dyes listed above can be used at a very low concentration the lowest concentration consistent with a reasonable signal to noise and that can be imaged with low light intensities.

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

381

This is necessary because fluorescence from the dyes becomes a non-linear function of dye concentration as the concentration increases, and also because all of these indicators act as photosensitising agents and are associated with photodynamic injury to cells, culminating in mitochondrial depolarisation and ATP depletion. For TMRM or TMRE in our hands the optimal concentration is
10–40 nM, but this may vary with cell type, perhaps due to variability of the plasma membrane potential. The approach depends on the principle that, at very low concentrations, the fluorescence signal shows a linear relationship with dye concentration and the dye concentration should simply reflect the Nernstian distribution of dye between compartments in response to local potential differences. According to this principle, one would expect an approximately 10 fold concentration of dye from saline to cytosol (given a plasma membrane potential of )60 mV) and then a concentration of some 400–800 fold from the cytosol into the mitochondria if Dwm lies at about )150 to )180 mV. Thus, the fluorescence signals may range in intensity some 3–4000 fold from the bathing saline to mitochondria, so that resolution of the signal in both cytosol and mitochondria by any imaging system demands digitisation of the signal to at least 12 bits (4096 grey levels). The concentrating effect of the plasma membrane potential is very important, and changes in this value will have a major impact on the accumulation of dyes by the mitochondria––a variable ignored by most in the field. It is terribly important to bear in mind that many manipulations in which people are interested in measuring mitochondrial potential––often related to cell death or injury––will also be associated with alterations in plasma membrane potential, and appropriate experimental controls must be made. The removal of contaminating out of focus signal is also necessary if measurements are to be made accurately. Even with reasonably high resolution, good quality imaging on a cooled CCD camera, the loss of spatial resolution caused by out of focus signal makes it impossible to measure discrete mitochondrial signals accurately, except perhaps in very thin cells (e.g. see Duchen et al., 2001). These issues have been discussed very carefully by Fink et al. (1998). For experiments where dyes are used in the redistribution mode, cells are bathed in the dye at very low concentrations (10–30 nM) and allowed to reach equilibrium. This may take 30 min or more, and the time required must be established to ensure that a steady state has been reached before measurements are made. For all experimental work, the dye must be present in all bathing solutions. The approach is effectively used in two different types of experiment: (a) dynamic measurements in which one asks how the potential will change in response to some change in the cell––cell signalling, or some pathophysiological condition, or (b) a comparison of populations of cells that have been previously exposed to different conditions. When dynamic measurements are made, a fall in mitochondrial potential promotes the redistribution of indicator between compartments in response to mitochondrial depolarisation, dye will leave mitochondria and move into the cytosol. In the short

382

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

Fig. 4. Measurements of mitochondrial potential. (A) These images show the distribution of TMRM in an astrocyte before (i) and after (ii) application of the uncoupler FCCP to dissipate mitochondrial potential. Cells were bathed in 40 nM TMRM which was allowed to equilibrate over 30 min. The dye accumulates in response to the membrane potentials in each compartment of the cell and is clearly concentrated into mitochondria. The series of images shown in (B) illustrates the dequench of Rhodamine 123 following mitochondrial depolarisation. A culture of astrocytes was loaded with rh 123 at a high concentration (
2 lM) for 15 min and then washed. The image sequence has been processed so that the whole sequence was divided by the first image in order to show the relative change in signal over each pixel. Upon application of FCCP, there are immediate local increases in signal around the mitochondria as the dye is released from the mitochondria into the cytosol. This increases with time and the dye gradually accumulates in the nucleus where the cell is at its thickest. The mitochondria are left as dark shadows where the signal has changed little.

term, this can be resolved as a decrease in the specific mitochondrial signal and an increase in signal over cytosol and/or the nucleus (Fig. 4A). In the short term (10 s of

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

383

seconds), this may not be associated with any change in signal at all over the whole cell and, if repolarisation is rapid, the dye will simply be taken back up into mitochondria (see, for example, Huser and Blatter, 1999). However, if the mitochondrial depolarisation is sustained, then the loss of mitochondrial dye to the cytosol will be followed inevitably by a re-equilibration of the cytosolic dye fraction, which will now move to the bathing saline, causing a loss of signal from the whole cell. Similarly, an increase in mitochondrial potential will initially cause an increase in specific mitochondrial signal and a decrease in the cytosolic signal with no change in net signal over the cell until the cytosol re-equilibrates with the bathing saline, and the mean signal from the cell then increases. In the second model, in which populations of cells are compared, the assumption is made that at equilibrium, the mitochondrial fluorescence signal will reflect the mitochondrial dye concentration, which is a direct function of the potential. Thus, if a manipulation has caused a loss of mitochondrial potential, the mitochondria will accumulate less dye and the mean signal from the mitochondria will be reduced at a steady state. Confocal imaging with fixed confocal optical thickness, laser power and detector sensitivity should allow quantitative comparisons to be made using these non-ratiometric indicators (e.g. see Beltran et al., 2000), but with some important caveats, which we discuss below. It is also worth considering using cell sorting techniques for such measurements, but there are some real problems which I will highlight below, and so we would recommend a combination with confocal imaging to ensure the security of the data if at all possible. 3.1.2. Measurement of Dwm (ii) the quench/dequench mode It should be clear, that while the approach outlined above is theoretically rigorous, it is also fraught with potential misinterpretations and errors. These problems have been discussed at some length by Ward et al. (Ward et al., 2000), and by Rottenberg and Wu (Rottenberg and Wu, 1998). When we started this kind of work in the late 1980s, we were obliged to use photomultiplier tubes to measure the averaged signal across a cell. In this case, if a dye simply redistributes from the mitochondria to the cytosol in response to a mitochondrial depolarisation, the mean signal may not change at all. Several groups therefore adopted a strategy which has certainly confused many: dyes such as TMRM, TMRE and rhodamine 123 were loaded into cells at relatively high concentrations (
1–20 lM) for short periods–– usually 10–15 min––followed by washing. As the dyes accumulate into the mitochondria, they reach concentrations at which the signal shows a phenomenon called autoquenching––energy is transferred by collisions between monomeric dye molecules and the concentration of dye may promote formation of aggregates of dye molecules which may be non-fluorescent. Redistribution of the dye into the cytosol as the mitochondria depolarise relieves the quench and the net fluorescence signal increases. Thus, in this mode, mitochondrial depolarisation is associated with an increase in fluorescence. This is illustrated in Fig. 4B, in which I have divided a series of images by the first of the sequence. In this way, only relative changes in signal are seen over each pixel. The series shows the change in fluorescence from an astrocyte loaded with rh 123 and then exposed briefly to FCCP. Note that puffs of bright

384

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

signal are seen first around the mitochondria and then the cytosol becomes progressively brighter with the dye gradually accumulating in the nucleus, which has a bigger volume in these cells than the very flat cytosol. The mitochondria now appear as dark shadows against this bright background as the signal within the mitochondria has if anything decreased, while the signal throughout the cell has increased dramatically––typically a change of 2–3 fold. In our hands, these signals have always behaved exactly as predicted from chemiosmotic theory, and the signal provides an apparently reliable way to follow changes in Dwm with time. Another important point here is that measurements of the mean signal from a cell will give information about the average change in mitochondrial potential in that cell, and so useful information can be obtained without demanding high resolution imaging. Indeed, we have routinely exploited this approach in low power imaging of a field of cells in which we can bathe the cells with a given drug and follow changes in Dwm in 20–30 neurons with time (e.g. see Vergun et al., 1999). It also turns out that rhodamine 123 is significantly less permeant across membranes than TMRM or TMRE (see Bunting, 1992; Ward et al., 2000). This has the result that the redistribution across the plasmamembrane in response to depolarisation of Dwp is much slower than in the case of TMRM, and so, even though the approach might seem less rigorous than the (re)-distribution approach defined above, the dequench signals obtained with rhodamine 123 seem to be more reliable and unambiguous indicators of changing Dwm and are less affected by changing Dwp than any other that we have used. It is important to understand that the approach is completely useless in attempting to compare mitochondrial potentials in populations of cells in response to manipulations, as the fluorescence signal is a non-linear function of dye concentration, and so absolute intensity is now meaningless. 3.2. Assessing redox state in situ––measurements of NAD(P)H and flavoprotein autofluorescence The term autofluorescence refers to the fluorescence that arises from endogenous compounds intrinsic to the cell, and is used to distinguish it from the fluorescence of indicators that are artificially introduced. Our understanding of these signals and their properties owes much to the pioneering work of Britton Chance in the 1950s and 60s, who showed that the bulk of intrinsic fluorescence in most cells and tissues arises from NADH and flavoproteins, the bulk of which is mitochondrial, although many cells also show some non-mitochondrial green/yellow fluorescence whose origin seems less clear. These signals can provide valuable indicators of changes in mitochondrial metabolism, as their properties change with the redox state of the carriers. Thus the fluorescence of NADH is excited in the UV (peak excitation at about 350 nm) and emits blue fluorescence (with a peak at about 450 nm, see Chance et al., 1979; Eng et al., 1989). The oxidised form, NADþ , is not fluorescent. An increase in UV-induced blue fluorescence therefore indicates an increase in the ratio of NADH to NADþ ––a net shift in the pyridine nucleotide pool to the reduced state.

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

385

These changes in signal do not indicate net changes in the absolute size of the total pool but rather a change in the balance of reduced to oxidised forms. NADPH is also fluorescent with very similar spectral properties, but in most cell types, seems to be present at much lower concentrations than NADH. NADH and NADPH are both present in both mitochondrial and cytosolic compartments, however several properties tend to mitigate in favour of the mitochondrial signal––there is more of it, and the binding of NADH to membranes enhances the fluorescence while enzymatic binding is thought to quench the cytosolic fraction (Chance and Baltscheffsky, 1958). Flavoproteins ferry electrons using a flavin or FAD molecule. Flavoprotein fluorescence is excited in the blue (with a peak at about 450 nm) and fluorescence emission is maximal in the yellow/green, with a peak at about 550 nm. In contrast to NADH, flavoprotein fluorescence decreases when the carrier binds electrons. A decrease in flavoprotein autofluorescence reflects an increase in the ratio of reduced to oxidised flavoprotein––the inverse of the response of the pyridine nucleotides. The redox state of the pyridine nucleotide and flavoprotein pools reflect the balance between the rate of reduction by substrate processing and the rate of oxidation by mitochondrial respiration. Thus, both upregulation of substrate processing and inhibition of respiration will favour a net balance towards a reduced state. Conversely, an increase in respiratory rate favours net oxidation of the pool. Thus, upon inhibition of respiration by CN– or by anoxia, the respiratory chain cannot oxidise the reduced forms which will accumulate to a new steady state. NADH autofluorescence increases and flavoprotein autofluorescence falls (Fig. 5A and B). Mitochondrial respiration responds to collapse of Dwm by an uncoupler with an increase in respiratory rate. This promotes maximal oxidation of NADH to NADþ and FADH2 to FAD, decreasing the autofluorescence from NADH and increasing that from FAD (Fig. 5A and B). Cardiac myocytes show a very bright and robust autofluorescecnce signal, presumably in part because they are so tightly packed with mitochondria. In the UV-confocal images shown in Fig. 5A, note how the UV elicited autofluorescence is clearly localised to mitochondria and effectively disappears in response to oxidation after application of FCCP (Fig. 5C). Experimentally, we have observed that in some cells the UV elicited autofluorescence is uniquely mitochondrial, while other cells show some other fluorescent structures––in some adipocytes, we have seen some blue non-mitochondrial fluorescence in fat droplets for which we cannot at present account. Similarly, in some cell types, we can see almost no blue autofluorescence at rest but the mitochondria appear to emerge out of a background when respiration is stopped with cyanide, suggesting that in these cells NADH is maximal oxidised at rest. A systematic examination of the maximally reduced and maximally oxidised levels of both NADH and FAD may provide unique information about respiratory activity at the level of the single cell. The application of this technology has proven enormously useful to us, as it gives a way to assess redox state at the level of single cells where changes can be related to cell type or to other aspects of cell specialisation. We recently used this approach to ensure the absence of respiration from individual cells that we had made free of mitochondrial DNA––q0 . In these cells, which have no respiratory chain, cyanide and FCCP had absolutely no effect at all on mitochondrial NADH autofluorescence,

386

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

Fig. 5. Measurements of redox state. (A) Assessing redox state in an adipocyte cell line using cyanide to promote maximal reduction of flavoproteins and NADH and FCCP to promote maximal oxidation. The confocal system allows sequential switching between excitation at 351 nm (for NAD(P)H measured between 420 and 480 nm) and 458 nm (for flavoproteins, measured at >505 nm). The changing intensities of the two signals are plotted in (B), normalised between the maximally reduced (CN) and maximally oxidised (FCCP) signals. In (C) are shown images of NAD(P)H fluorescence from a freshly isolated rat cardiomyocyte before (i) and after (ii) application of FCCP. These images and those below were obtained on a UV-vis confocal system. When illuminated with low level UV light, the mitochondrial localisation of the resulting intrinsic blue fluorescence is obvious. Application of FCCP depolarises the mitochondria stimulating respiration and therefore rapidly oxidising the NAD(P)H to the non-fluorescent form, NADþ (ii). What is left is the cytosolic signal from NADH and NADPH which does not change acutely in response to FCCP.

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

387

and CN– had no effect on mitochondrial potential, but FCCP collapsed the small potential that these mitochondria maintain. Clearly, then measurement of these signals alone do not provide useful or meaningful measurements, but changes in signal may give information about changing substrate utilisation or oxygen consumption which is otherwise inaccessible, especially if calibrated against the maximal reduced level (achieved by complete inhibition of respiration with cyanide) and the maximally oxidised level (achieved by maximal stimulation of respiration with an uncoupler such as FCCP). 3.3. Location of proteins to mitochondria––using fluorescent mitochondrially targetted proteins There remain several significant problems with fluorescence imaging approaches to measurement of intramitochondrial calcium. Most particularly at present, none of the available indicators is sufficiently selective in localising to mitochondria, and so signals are always contaminated by cytosolic signal (see above). In addition, most of the available indicators that do target to mitochondria have a relatively high affinity for calcium and so are likely to be saturated when intramitochondrial calcium concentration is high. Direct and accurate quantitative measurements of intramitochondrial calcium is probably best achieved by transfection of cells with the Ca2þ -sensitive protein probes targetted to the mitochondrial matrix, a technique originally pioneered by Rizzuto and his colleagues in the early 1990s. This group introduced the use of the photoprotein Aequorin to measure calcium concentrations in specific compartments in cells. Cells are transfected with a chimeric cDNA encoding aequorin and a mitochondrial presequence from the mitochondrial enzyme cytochrome oxidase in order to target the fusion protein specifically to mitochondria (Rizzuto et al., 1993, 1994). This is an elegant technique ensuring specific measurement of Ca2þ changes within a discrete compartment. The signal generated by the photoprotein expressed this way is generally too weak to be useful for [Ca2þ ]m imaging at the single cell level using current technology, and data are therefore gathered from a population of cells. Aequorins are available that have been engineered with lower affinity for calcium, allowing some fascinating insights into mitochondrial calcium handling. The approaches discussed above are nicely complementary with the use of recombinant aequorin; spatial information is inferred from the selective expression of the reporter protein in different compartments in the cells, and while single cell imaging using fluorescent dyes have limitations of accurate compartmentalisation, the approach provides details of the spatiotemporal features of the signalling pathways. More recently brighter fluorescent targeted protein probes have been developed that promise to combine the benefits of specific targeting and manipulation possible with protein based probes with the high resolution imaging made possible with (relatively) bright fluorescence signals. These probes are currently represented by the cameleons and pericams (Nagai et al., 2001), probes based on the fluorescence of the green fluorescent protein (GFP) of the jellyfish Aequorea victoria,. In each case,

388

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

the protein has been engineered with a calcium binding domain of calmodulin. The pericam uses a circular mutated GFP linked with calmodulin and it’s target protein, M13, so that calcium binding changes conformation of the protein and changes the fluorescence intensity. The cameleons (Miyawaki et al., 1999) also use a calcium binding site of calmodulin but signal the change in conformation on calcium binding by measuring the transfer of energy from a CFP to a YFP, a process of FRET (fluorescence resonance energy transfer) theoretically at least improving the signal/ noise. Tagging with fluorescent adducts has provided a wealth of information on the distribution and trafficking of a variety of intracellular proteins (for review see Cubitt et al., 1995). Such use of the green fluorescent protein (GFP), derived from the jellyfish Aequorea victoria, is now widespread, and in combination with confocal microscopy has provided new insights into mitochondrial protein movements and distribution. Moreover, genetically modified mutants of GFP have provided probes with improved spectral properties greatly enhancing the usefulness of the fluorophores. GFPs with more than 20-fold stronger fluorescence than wild-type have been produced by mutagenesis and selection of the products by FACS analysis (Cormack et al., 1996). Additionally, thermotolerant mutations of GFP have facilitated use of the protein in animal cells as the wild-type protein fluoresces poorly at temperatures over 25C (Siemering et al., 1996), and fluorescent proteins with altered spectral properties have been used in experiments using fluorescence energy transfer (FRET) (Mahajan et al., 1998). The 238 amino acid, 27 kDa protein may be attached to either the C- or N-termini of proteins (Wang and Hazelrigg, 1994) and due to its relatively compact nature retains the ability to diffuse throughout the cytosol and enter the nucleus. Several groups have utilised GFPs attached to mitochondrial targetting sequences in order to direct the fluorophore to mitochondria. The great majority of mitochondrial proteins are synthesised in the cytosol and much work has identified a host of protein presequences and other targetting sequences that direct proteins to the mitochondrial matrix, intermembrane space and inner or outer membranes (for extensive review seeNeupert, 1997). By using known targetting sequences spliced to the cDNA coding for GFP, the GFP protein can be expressed in vivo and then targetted to the mitochondrial region of interest. Rizutto and colleagues have shown how the targeting presequence of human cytochrome c oxidase (COX, Complex IV of the respiratory chain) can be attached to photoproteins (initially aequorin, however the same technique may be used with GFPs) to allow mitochondrial targeting and important of the functional protein into the mitochondrial matrix (Rizzuto et al., 1992). After digestion with restriction enzymes to excise and isolate the targetting sequence, as well as and the cDNA encoding a few of the amino acids of mature COX, the resultant cDNA fragment can be ligated to the cDNA encoding the GFP. By including a portion of the mature COX, the authentic cleavage site of the targeting sequence is retained and then cleaved from the mature photoprotein. Such a technique may be used to examine the distribution of mitochondria in cells where the use of fluorescent dyes would be impractical. For example, we have used a mitochondrially targeted GFP to establish the location of mitochondria in fixed astrocytes where the fixation process induces redistribution of cationic fluorophores.

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

389

Although the targeting sequence for COX is widely used, alternative presequences may be spliced to the GFP cDNA. Thus markers can be directed highly specifically to their subcellular targets allowing, for example, characterisation of the signal that directs the Tom 20 protein to the outer membrane (Kanaji et al., 2000), and localisation of the key enzyme carnitine palmitoyltransferase 1 to the cytosolic face of the outer mitochondrial membrane (van-der-Leij et al., 1999).

4. Mitochondrial function and cell signalling 4.1. Mitochondrial calcium uptake While the principle function of the mitochondrial potential is clearly to drive ATP synthesis, it also provides the major mechanism to handle calcium, and I will dwell on this process, partly because it has been quite controversial over recent years but more particularly because it seems ever clearer that mitochondrial calcium handling is strongly implicated in many disease processes, as the accumulation of calcium into mitochondria has major functional consequences for the mitochondria and for the cell both in terms of physiology and pathology. For many years there was some doubt and dispute about the physiological significance of mitochondrial calcium uptake, but now it seems indisbutable that most physiological calcium signals in most cells will be associated with some measure of calcium accumulation in mitochondria (for a recent review see Rizzuto et al., 2004). Furthermore, it is now widely held that many of the pathological states in which mitochondrial dysfunction has been implicated may involve alterations in mitochondrial calcium handling or a pathological effect of mitochondrial calcium ‘overload’ (see below). It is therefore crucial to understand the basic functional machinery involved in this pathway. Mitochondrial calcium uptake is primarily driven by the electrochemical potential gradient established by the mitochondrial membrane potential and by a relatively low intramitochondrial Ca2þ concentration [Ca2þ ]m . Whenever energised mitochondria are exposed to raised [Ca2þ ]c , Ca2þ will move from the local environment into the matrix, driven by the electrochemical potential gradient for Ca2þ . Thus, Ca2þ movement into mitochondria depends on the mitochondrial membrane potential and also on the intramitochondrial [Ca2þ ] ([Ca2þ ]m ), which is generally believed to be kept low under ‘resting’ conditions largely through the activity of a xNaþ /Ca2þ exchanger (see below). Typically, mitochondrial calcium uptake follows the cytosolic signal with a time lag and re-equilibration of mitochondrial calcium is realtively slow so that mitochondria may effectively seem to integrate cytsolic calcium signals over time (see Fig. 6 and Hajnoczky et al., 1995). Ca2þ is taken up through the mitochondrial inner membrane by a uniporter. The biophysical properties of this pathway and its molecular identity have yet to be defined. Thus, it is not yet known with certainty whether it is a channel or a carrier. Flux rates are similar to those measured for fast gated pores, but are slower than those seen for most classes of ion channel (Gunter et al., 2000, for review). Thus, uniporter activity shows little sensitivity to changes in temperature, and also shows a

390

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

Fig. 6. Some characteristics of mitochondrial calcium uptake. (A) Dual loading of cells with a primarily cytosolic [Ca2þ ]c and a primarily [Ca2þ ]m indicator allows direct measurement of the relationship between the two functions. In the example shown (modified from Rakhit et al. (2001)) neonatal cardiomyocytes were loaded with fluo-4 and with rhod-2. The change in mitochondrial signal in the rhod-2 signal was clearly seen after depolarisation with 50 mM KCl (before––ii, and after––iii). In panel iv the cytosolic signal and mitochondrial signals are plotted and the difference in time course between the two is clear. (B) A similar experiment was carried out in a HeLa cell in response to the application of ATP which mobilises Ca2þ from ER through an action at P2 receptors. The cytosolic signal shows a characteristic oscillation while the mitochondrial signal shows the gradual integration of the signal as recovery is too slow to follow the oscillation rate.

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

391

wide spectrum of cation selectivity (Ca2þ > Sr2þ > Ba2þ > Mn2þ > Mg2þ ), more consistent with the activity of a channel than a carrier. Ca2þ uptake via the uniporter is inhibited by ruthenium red (RuR), which inhibits a variety of different classes of cation channels, including L-type plasmalemmal Ca2þ channels (Duchen, 1992), ryanodine sensitive ER Ca2þ release channels (Lukyanenko et al., 2000), and vanilloid receptor operated channels (Wood et al., 1998), again suggesting that the uniporter is likely to exhibit the properties of an ion channel. The pathway is more selectively inhibited by Ru360 and by diamino-pentane pentammic acid (DAPPAC, Crompton and Andreeva, 1994). DAPPAC is impermeable through the plasma membrane while Ru360 may be partially permeable. Very recently, patch clamp recordings from mitoplasts have revealed a Ca2þ selective inwardly rectifying channel/current in the inner mitochondrial membrane which has characteristics consistent with the properties of the uniporter (Kirichok et al., 2004), although the molecular identity of the channel remains to be identified. One of the most striking features of the conductance described in these studies is the strong inward rectification that would make mitochondrial Ca2þ accumulation very voltage sensitive, and that would also strongly limit the reversal of the pathway to allow calcium efflux from mitochondria at depolarised potentials. The latter seems odd, in a way, as an increase in cytosolic [Ca2þ ] induced by mitochondrial depolarisation with an uncoupler has long been a standard way to try to assess mitochondrial calcium content, but this may be dependent on other ambient conditions in the cytosol that were not reproduced by the experimental design in the patch studies. It is also possible that the calcium efflux routinely seen in this way is not carried through the uniporter. The very steep voltage sensitivity of mitochondrial calcium import has been described for some time (see Kapus et al., 1991) and has interesting functional implications. Thus, even modest mitochondrial depolarisation could have quite a profound impact on mitochondrial calcium uptake, and this may be important both in terms of the consequences for physiogical calcium signalling and in terms of pathophysiology (see below). In this way, mechanisms that cause small changes in mitochondrial potential may be surprisingly cytoprotective in pathological conditions of calcium overload (see below and Rakhit et al., 2001). Another remarkable feature of the uniporter is an apparent gating by [Ca2þ ]c , identified primarily through studies of the Ca2þ sensitivity of RuR sensitive mitochondrial Ca2þ efflux in response to dissipation of Dwm (Igbavboa and Pfeiffer, 1988; Montero et al., 2001). In a recent study, Montero et al. (2001), showed that, although collapse of Dwm prevents mitochondrial Ca2þ uptake, collapse of Dwm after the accumulation of mitochondrial Ca2þ inhibited mitochondrial efflux, i.e. all mitochondrial efflux pathways were inhibited by depolarisation. Addition of Ca2þ to the depolarised, Ca2þ loaded mitochondria then promoted mitochondrial Ca2þ release sensitive to RuR, suggesting release through the uniporter. This is consistent with suggestions that the uniporter is allosterically gated by [Ca2þ ]o (Igbavboa and Pfeiffer, 1988), an observation that may also explain why local [Ca2þ ]c needs to be higher than one might expect simply from the behaviour of a conducting Ca2þ channel in order to see significant increases in [Ca2þ ]m .

392

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

An uptake pathway with properties distinct from those of the uniporter has also been described (Sparagna et al., 1995; Buntinas et al., 2001). This has been referred to as the rapid uptake mode (RaM). This pathway has the capacity to transfer Ca2þ very rapidly into the mitochondria during the rising phase of a Ca2þ pulse. The properties of the pathway differ in different tissues (Buntinas et al., 2001), but in heart, the pathway saturates quickly and is slow to reset after activation. Again, the functional significance of the pathway remains to be established. It has also been suggested that the ryanodine receptor, the channel of calcium induced calcium release in cardiac muscle, may also be expressed in mitochondria and may represent a calcium influx pathway into mitochondria (Beutner et al., 2001). This has since been identified as the RyR1 isoform of the RyR. The evidence was compelling. The authors stopped short at suggesting that this protein is the uniporter, especially as it appears only to be found in excitable tissues––heart and brain––while the uniporter is ubiquitous. The uniporter certainly shares a remarkable array of properties with the RyR, and the study raises a number of interesting issues. For example, is calcium uptake into mitochondria an equivalent process in all tissues or is it feasible that different tissues express different proteins so that the details of mitochondrial calcium handling may show significant differences between tissues? Almost all the information we have about the behaviour of the uptake pathways comes from studies of isolated mitochondria which tend perforce to come from tissues in which it is possible to obtain a substantial yield––the heart, whole brain, and liver. We know relatively little about specific properties of the pathways in specific cell types in which it is not feasible to obtain pure preparations of isolated mitochondria. Hence the importance of progress in studies which allow identification of pathways at the level of the single cell. Another important potential consequence of this discovery is that, as there are recognised mutations of the RyR that cause disease, perhaps such mutations might also involve mitochondrial calcium handling pathways. Further, there are also important regulatory proteins associated with the RyR. Does this also raise the possibility that there might be similar regulatory proteins associated with the uniporter? A recent study by Pacher and Hajnoczky (2001) reported that Ryanodine had no effect on mitochondrial calcium uptake in two cell lines, rbl and H9C2. The latter is an excitable cell line derived from muscle. The issue undoubtedly remains controversial, but like so many such findings, it may provoke interesting discovery by those who choose to investigate further. One fascinating area that still seems open is the issue of modulation or regulation of uniporter activity. Are there physiological or pathophysiological signals which vary mitochondrial Ca2þ uptake either to increase or decrease calcium accumulation in different contexts of cell signalling? It has recently been suggested that protein phosphorylation may modulate mitochondrial calcium accumulation, although the mechanism (i.e. the targets of phosphorylation) to my knowledge remain unknown. Thus, Montero et al., have recently shown that an inhibitor of p38 MAPkinase dramatically increased mitochondrial calcium accumulation compared to the cytosolic signal (Montero et al., 2002), suggesting that the tonic activity of the kinase must suppress or limit mitochondrial Ca2þ accumulation. Overexpression of some of

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

393

the PKC isoforms also seems to modify mitochondrial Ca2þ signals (Pinton and Rizzuto, personal communication) while inhibitors of PKC also increase mitochondrial Ca2þ uptake (Montero et al., 2003). The mitochondrial outer membrane has been assumed to be permeant to small ions and so has been largely neglected in considerations of mitochondrial Ca2þ handling. However, the outer membrane may play a more significant role in modulating access of Ca2þ to the uniporter through the selectivity filter of the voltagedependent anion channel (VDAC). It appears that VDAC is Ca2þ permeant and is regulated both by [Ca2þ ] and by RuR (Gincel et al., 2001). This raises questions about the extent to which the properties of the uptake pathway are defined by VDAC acting as a first filter. It is also tantalising that VDAC appears to be part of the mitochondrial permeability transition pore (mPTP; see below), a pore which is itself regulated by [Ca2þ ]m , as the mPTP provides a potential efflux pathway for Ca2þ , although the physiological relevance of this pathway is debated. Such studies point to the outer membrane as a significant permeability barrier that may itself be regulated. 4.2. Mitochondrial Ca2þ efflux The major route for Ca2þ efflux from mitochondria is a xNaþ /Ca2þ exchange. Identified about 20 years ago, it has a discrete pharmacology distinct from the plasmalemmal exchanger. The stoichiometry of the exchanger seems still to be controversial. Initially, it was thought to be an electroneutral 2Naþ /Ca2þ exchanger (Brand, 1985), but this has been questioned, as the exchanger can operate against a [Ca2þ ] gradient whose energy is over twice that of the Naþ gradient (Jung et al., 1995). Jung et al. (1995) suggested a stoichiometry of 3Naþ /Ca2þ , in which case the operation of the exchanger will be dependent on Dwm . The inhibition of mitochondrial Ca2þ efflux by mitochondrial depolarisation (see above Montero et al., 2001; Bernardi and Azzone, 1982) supports this electrogenic stoichiometry. An electrogenic stoichiometry also predicts that Ca2þ efflux should be associated with mitochondrial depolarisation. To my knowledge, this has not been documented. This is a significant issue, as the stoichiometry will define the way in which calcium handling changes with mitochondrial potential, and needs to be clarified. 4.3. Mitochondrial influence on resting cytosolic calcium concentration––the ‘set point’ Flux studies in isolated mitochondria revealed many years ago that mitochondria will take up Ca2þ . With small elevations of [Ca2þ ]o , the removal of Ca2þ from the matrix by the xNaþ /Ca2þ exchange may be sufficiently rapid that net [Ca2þ ]m changes little. An increase in [Ca2þ ]o above
4–500 nM will exceed the capacity of the exchanger and then mitochondria will show net accumulation of Ca2þ . This was termed the ‘set point’ for mitochondrial Ca2þ uptake by Nicholls and Crompton (1980). It is worth considering that Ca2þ flux into mitochondria is not necessarily synonymous with a net increase in [Ca2þ ]m , especially considering our ignorance of the properties of Ca2þ buffering by the matrix. This is not purely semantic, as Ca2þ

394

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

uptake by the uniporter is electrogenic and is therefore associated with small changes in Dwm . Experimentally, changes in Dwm will reflect the rate of Ca2þ flux, and may therefore prove a more sensitive measurement of Ca2þ movement into mitochondria than measurement of [Ca2þ ]m . Further, net mitochondrial Ca2þ accumulation will be partly set by the activity of the xNaþ /Ca2þ exchanger––and we still know little about its regulation. Many excitable cells respond to depolarisation with a rise in [Ca2þ ]c which rises rapidly and recovers with an initial rapid phase and a slower second phase that can even form a plateau (Thayer and Miller, 1990; Babcock et al., 1997). It has been established in many preparations that the slow recovery phase reflects the redistribution of mitochondrial Ca2þ through the activity of the Naþ /Ca2þ exchanger, reflecting the set point, typically initiated at a [Ca2þ ]c of
500 nM. The operation of this system has functional consequences at presynaptic terminals, where the [Ca2þ ]c plateau that follows repetitive stimulation, maintained by the re-equilibration of mitochondrial Ca2þ , provides an elevated [Ca2þ ]c baseline upon which subsequent stimulation initiates an enhanced synaptic response––the basis for post-tetanic potentiation of synaptic transmission David et al., 1998; Tang and Zucker (1997). It is also intriguing that the post stimulus plateau phase is not seen in non-excitable cells following the transmission of [Ca2þ ]c signals from ER to mitochondria. Certainly in astrocytes, [Ca2þ ]m remains high for a very prolonged period after stimulation (Boitier et al., 1999), suggesting that mitochondrial Ca2þ efflux must be very slow and perhaps the activity of the exchanger differs between tissues or cell types. 4.4. Mitochondria and [Ca2þ ]c microdomains There has been some debate about the quantitative relationships between ambient (cytosolic) [Ca2þ ] and mitochondrial uptake. In HeLa cells transfected with mitochondrially targetted aequorin, permeabilised, and then exposed to added [Ca2þ ] buffer solutions, net mitochondrial Ca2þ accumulation was only detectable if the added Ca2þ reached concentrations higher than 3 lM, while [Ca2þ ]c signals evoked by IP3 mobilising agonists were far more effective at raising [Ca2þ ]m even though the mean [Ca2þ ]c signal in the cytosol might rise to <1 lM (Rizzuto et al., 1993). This observation led to the suggestion that mitochondria must be positioned at privileged sites close to the ER Ca2þ release sites where they would be exposed to microdomains of high local [Ca2þ ]c sufficient to promote rapid Ca2þ uptake. Further studies from the same lab showed that mitochondria may form contact sites with ER to provide a morphological basis for the existence of such local microdomains, and there is good EM supporting evidence that mitochondria may form very close contacts with ER or SR. The proximity of mitochondria to SR or ER Ca2þ release sites has been further emphasised through evidence that focal, non-propagating ER/SR Ca2þ release can cause a transient increase in [Ca2þ ]m in mitochondria close to the release site. Thus, we found (Duchen et al., 1998) that mitochondria in cardiomyocytes show spontaneous transient mitochondrial depolarisations that were dependent on local SR Ca2þ release and were blocked by inhibition of mitochondrial Ca2þ uptake. Pacher et al.

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

395

(2002) in a series of elegant experiments have demonstrated directly the transfer of calcium from ER to mitochondria and have more recently shown that local [Ca2þ ]c sparks may be associated with the direct transfer of Ca2þ to mitochondria visualised as transient increases in [Ca2þ ]m ––which the group termed Ca2þ ‘marks’. Further data from cardiomyocytes (Sharma et al., 2000) strongly suggests that, in cardiomyocytes, mitochondria and SR must show very close coupling, as the transfer of Ca2þ to mitochondria in response to SR Ca2þ release with caffeine in permeabilised cells was sustained despite Ca2þ buffering by BAPTA sufficient to suppress the cytosolic signal. The transfer of Ca2þ to the mitochondria was prevented by disrupting the cytoskeleton, suggesting the maintained close apposition of mitochondria to SR was central to this signal. The proximity of mitochondria to Ca2þ release sites has functional consequences for [Ca2þ ]c signalling. Using [Ca2þ ] indicators in both mitochondria and ER in permeabilised cells, Csordas et al. (1999) showed direct transfer of Ca2þ from ER to mitochondria and suggested that the proximity must be
10–20 nm. This work was extended to show that mitochondrial Ca2þ uptake enhances the release of Ca2þ from the ER in response to IP3 by acting as a local buffer (Hajnoczky et al., 1999). Thus, by removing Ca2þ from the microdomain close to the IP3 Ca2þ release channel, mitochondria prevent the Ca2þ dependent inactivation of the channel and facilitate ER Ca2þ release. This mechanism allows mitochondria to play a significant role in shaping the spatiotemporal patterning of [Ca2þ ]c signals. In Xenopus oocytes, energisation of mitochondria enhances the propagation and coordination of [Ca2þ ]c waves (Jouaville et al., 1995) while in astrocytes, which express primarily IP3 type 3 receptors, energised mitochondria served as a spatial buffer which limit the rate and extent of propagation of [Ca2þ ]c waves (Boitier et al., 1999). The principle is that the release of Ca2þ from IP3 sensitive stores is itself strongly regulated by local [Ca2þ ], sensitised by moderately raised [Ca2þ ] and inhibited when [Ca2þ ] rises above about 106 M. Thus, by regulating local [Ca2þ ] the mitochondria regulate ER Ca2þ release in response to IP3. Microdomains of [Ca2þ ]c regulated by mitochondria also play a significant role in the regulation of capacitative Ca2þ influx (Hoth et al., 1997; Gilabert et al., 2001), suggesting that the mitochondria must be positioned close to the plasma membrane. The principle is very much as outlined above for the IP3 receptor, as the Ca2þ influx channel is desensitised by Ca2þ . By keeping [Ca2þ ]c low in microdomains close to the channels, mitochondria keep the channels open and facilitate Ca2þ influx through the channels. In the blowfly salivary gland and in pituitary gonadotropes, mitochondrial Ca2þ uptake may even play a major role in defining the rates of oscillation of the IP3 generated [Ca2þ ]c signal (Zimmermann, 2000; Kaftan et al., 2000), suggesting that the interplay between mitochondrial Ca2þ uptake and ER Ca2þ release contribute significantly to the temporal patterning of the [Ca2þ ]c signal. In pancreatic acinar cells, the mitochondria are concentrated into a band that isolates the secretory pole of these polarised cells, and they seem to act as a ‘firewall’ that limits the spread of [Ca2þ ]c signals from their initiation at the apical pole to the basal pole (Tinel et al., 1999). Furthermore, mitochondria localised close to the basal

396

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

pole are more sensitive to local Ca2þ influx by capacitative entry, and so it seems that the positions of mitochondria within the cell may have a profound influence on their interaction with cellular [Ca2þ ]c signals (Park et al., 2001). This issue alone is fascinating but unresolved––what dictates the positions mitochondria occupy within cells? Indeed, imaging mitochondria within cells shows that they move, and that the movement is erratic and unpredictable, and, to my knowledge, we have no solid foundation to understand the functional significance of that movement, although it has been suggested that mitochondrial movement is at least in part dictated by energetic requirements (see Hollenbeck, 1996). It is also not yet entirely clear whether the maintenance of mitochondrial ER microdomains has a structural basis. In other words, are mitochondria specifically localised to be close to ER by some targeted, organised process and through the specific activity of tethering proteins, or are they simply close together because they are squashed together in a limited space? A recent study by Filippin et al. (2003), gave strong support to the former view, showing that mitochondrial––ER junctions at which Ca2þ transfer was most efficient were highly stable in space, despite the constant movement of both ER and mitochondrial structures. As far as I am aware, we currently have no idea what mechanisms might be involved in tethering or maintaining contact sights between ER and mitochondria or between mitochondria and the plasma membrane. Are [Ca2þ ]c microdomains essential for mitochondria to sense changes in [Ca2þ ]c associated with [Ca2þ ]c signals? In permeabilised adrenal glomerulosa cells, Szabadkai et al. (2001) found that graded additions of buffered external Ca2þ caused a graded but non-linear increase in [Ca2þ ]m , showing a response even when the ambient [Ca2þ ] was only
2–300 nM. Such data suggest that the close juxtaposition of mitochondria to Ca2þ sources is not an absolute requirement if they are to respond to [Ca2þ ]c signals. A recent study in HeLa cells (Collins et al., 2002) also suggests that areas of maximal mitochondrial Ca2þ uptake may be divorced from areas of maximal proximity with ER in HELA cells. This study suggested that peripheral mitochondria had larger mitochondrial potentials and that this might provide a mechanism to enhance mitochondrial [Ca2þ ]c accumulation into that mitochondrial population. The notion that mitochondria within a single cell may have different potentials remains contentious and the observation is critically dependent on the behaviour of the fluorescent indicators used to measure Dwm . This is probably not the place for further discussion of this issue, but my own view is that the question remains open and has not been satisfactorily resolved either way. An important issue in calcium signalling is simply the local availability of ATP to power the accumulation of calcium into SR/ER. Our own recent evidence (Dumollard, Carroll, Duchen, unpublished) suggests that in the mammalian egg, mitochondrial localisation close to ER may be critical in providing a local supply of ATP. The egg has a strange and very unusual metabolism, being almost exclusively dependent on mitochondrial oxidative phosphorylation, while glycolysis seems not to play any significant role in ATP supply and may even be harmful. In the egg, development is crucially dependent on coordinate calcium signals initiated by fertilisation, and any disruption of the shaping of the calcium signal will have drastic

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

397

consequences for development. The close dependence of ER Calcium loading on mitochondrial ATP supply means that any defect in mitochondrial function may have a disproportionate impact on subsequent development. 4.5. Mitochondrial Ca2þ uptake and mitochondrial function In teleological terms, it seems likely that the major functional significance of mitochondrial Ca2þ uptake is in the regulation of mitochondrial metabolism. In the early 1990s it became clear that the three major rate limiting enzymes of the citric acid cycle are all upregulated by Ca2þ (McCormack et al., 1990). The question that remained was the functional issue––is mitochondrial Ca2þ uptake sufficient for this mechanism to provide a functional regulation of metabolism? It seems now that the answer is an unequivocal ‘yes’. Changes in mitochondrial metabolism in response to changes in [Ca2þ ]c were demonstrated in 1992 (Duchen, 1992; Pralong et al., 1992) by measuring changes in mitochondrial NADH and flavoprotein autofluorescence in response to calcium signals in a variety of cell types. These observations showed clearly that (i) mitochondria must be taking up Ca2þ during [Ca2þ ]c signals, and (ii) that this was sufficient to activate the TCA cycle, causing increased net reduction of the cofactors. More recently, cells were transfected with firefly luciferase showing clearly and directly that mitochondrial Ca2þ uptake increases mitochondrial ATP production (Jouaville et al., 1999). The relative importance of this mechanism in the regulation of mitochondrial oxidative phosphorylation over the regulation of the rate of ATP generation by the ATP/ADP ratio is not clear. Attempts to measure the ATP/ADP ratio in relation to changes in work have always proved very difficult, but in general tend to show very little change and the ratio seems generally highly conserved. It therefore seems rather likely that the transfer of Ca2þ from the cytosol to mitochondria during [Ca2þ ]c signals represents a major mechanism to couple increased ATP demand with an increase in the supply, as in almost all systems, increases in work are associated with increases in [Ca2þ ]c (although this remains contentious: see Horikawa et al., 1998). Some very recent studies have added yet another step to this modulatory pathway. It seems that the glutamate/aspartate carrier, responsible for the transport of mitochondrial substrate, is also upregulated by calcium, and in cells overexpressing the carrier, stimulation of calcium signals was shown clearly to increase the rate of production of ATP (Lasorsa et al., 2003). The time course of the changes in [Ca2þ ]m and in activation of the enzyme systems becomes crucial. [Ca2þ ]c signals are typically brief, transient phenomena. Typically, it seems that the resultant mitochondrial activation is prolonged with respect to the change in [Ca2þ ]c (e.g. see Duchen, 1992; Hajnoczky et al., 1995; Robb-Gaspers et al., 1998). The failure of such mechanisms may also have profound functional consequences. For example, it has been shown that in cybrid cells containing mitochondria carrying a tRNALys mutation associated with the mitochondrial disease known as MERRF (myoclonic epilepsy with ragged-red fibers), mitochondrial calcium uptake was significantly impaired (possibly because the mitochondria may be depolarized, although

398

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

this was not established; see Brini et al., 1999). This will mean that mitochondrial oxidative phosphorylation cannot be upregulated in these patients in response to calcium signals and may contribute to symptoms of weakness in response, say, to exercise. Taken together, these studies suggest that a fundamental feature of the efficiency of transfer of the Ca2þ signal from ER to mitochondria is a function of the morphological characteristics of mitochondria in the cell particularly in relation to the sources and sinks of calcium. This leads to fascinating questions about the physiological regulation of mitochondrial structure and function. Under pathological conditions, mitochondrial Ca2þ accumulation may also play a key role in determining the outcome, acting as a trigger to pathophysiological events that may dictate the death of the cell. This will be discussed below.

5. Mitochondria and free radical generation In most cell types, mitochondria appear to represent one of the major sources of generation of free radicals or reactive oxygen species (ROS). This is because a consequence of the action of oxidative phosphorylation is the generation of unpaired electrons. The interaction of these electrons with O2 results in the generation of superoxide ions, highly reactive free radical species. These are readily interconverted to other radical species, such as hydroxyl ions (OH– ) and H2 O2 . It is widely assumed that the major impact of such ROS generation is harmful although recently it has been suggested that mitochondrial ROS generation may play a significant physiological signalling role. Nevertheless, ROS may cause lipid peroxidation and damage to cell membranes and to DNA, so that mitochondria represent not only a major source of ROS generation, but also a major target of ROS induced damage. It is worth remembering that this includes mtDNA, which has no associated histones and is less protected from radical damage than nuclear DNA. Seeing I have some modest freedom here to ramble, I have to say that I worry that the involvement of mitochondria and free radical generation in pathophysiology now seems almost to have become a mantra, repeated endlessly and too often without sound experimental basis. Nevertheless, there is no question that mitochondria are a major source of oxygen free radical (or reactive oxygen species, ROS) generation in most cell types, except perhaps for immune cells: macrophages, microglia in the CNS and neutrophils, which express an NADPH Oxidase and generate massive amounts of ROS after stimulation. Mitochondria provide an environment which depends on the transfer of electrons to oxygen, and so it is hardly surprising that at various stages in this process, singlet oxygen or superoxide may escape. Mitochondria are equipped with an armamentarium of antioxidant defences. They contain a high concentration of glutathione, a variant of superoxide dismutase (CuZnSOD) and catalase that removes the potentially harmful peroxide that is produced by the SOD. The specific sites of free radical generation along the electron transport chain seem somewhat controversial. Thus, in some tissues it seems that inhibition of electron transfer at

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

399

complex I (typically with rotenone), may generate an increase in radical formation, while in others, rotenone appears to reduce radical generation, presumably by preventing the passage of electrons further into the distal chain where the ROS are produced. It is widely accepted that a major site of radical generation lies at complex III. Methodological problems may underlie such differences, but in one study in which the same techniques were used side by side in mitochondria from kidney and from pulmonary vascular smooth muscle cells, the two populations of mitochondria responded in opposing ways to rotenone (Michelakis et al., 2002). The basis for such a difference remains obscure. Recent evidence suggests strongly that free radicals seem to be generated from complex I through oxidation of NADH on the matrix side of the inner mitochondrial membrane (St-Pierre et al., 2002). In intact mitochondria, superoxide generated in this way may be efficiently scavenged by intramitochondrial antioxidant defences, while superoxide production can be measured from submitochondrial particles or from mitoplasts in which such defences may be depleted (Chen et al., 2003). In general, the leak of electrons seems to be increased by an increase in mitochondrial potential and decreased with mitochondrial depolarisation (e.g. see Votyakova and Reynolds, 2001). An increase in potential slows respiration and permits electron leak at proximal parts of the chain. Inhibition of respiration distal to the point of radical generation (e.g. at complex IV with cyanide) is expected to increase ROS production, while uncouplers will reduce electron leak by increasing respiratory rate and collapsing the potential, favouring the transfer of electrons to oxygen at Complex IV and generating water. Recent studies also raise the suggestion that use of succinate as a main respiratory substrate can lead to radical generation by reverse electron flow from Complex II to complex I (via III; see St-Pierre et al., 2002). Nevertheless, much of this field remains poorly understood, highly controversial and full of misleading and confusing literature. This is partly because the methods available to measure rates of radical production from cells are insecure and prone to artefact and so very difficult to use to give absolutely secure and unambiguous data. Recent data from Adam-Vizi’s lab have suggested that a major source of oxyradicals may be the a-Ketoglutarate dehydrogenase itself. They have suggested that the limiting factor is the supply of NADH. Inhibition of respiration with rotenone will increase NADH and this in turn will encourage radical generation by the enzyme. This may be one reason why there has been controversy about the generation of radicals from complex I––perhaps some of these radical species are derived not from the respiratory chain but from the citric acid cycle intermediates (Adam-Vizi, personal communication). It has become a widespread belief that mitochondrial generation of ROS is a major cause of cellular damage that accumulates over time and could be responsible for aging. One recent television documentary tried to draw an analogy between aging and rusting iron girders, suggesting that both represent the harmful effects of oxygen. This is a highly controversial field and way beyond the scope of this review. It is certainly intriguing that several of the genes that seem to be important in aging seem to generate proteins that are somehow involved in responses to or control of oxidative stress (see Lin et al., 1998; Migliaccio et al., 1999; Fabrizio et al., 2001; Melov,

400

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

2002a,b), but it is important to recognise that this area is highly controversial and I will not deal with this further here. What is most interesting however is the recent suggestion that mitochondria have evolved mechanisms that regulate the rate of production of ROS through an intrinsic feedback loop involving the expression of uncoupling proteins. This is a field very much in its infancy, but worthy of comment.

6. Uncoupling proteins and mitochondrial oxidative stress The uncoupling proteins, or UCPs, are at present a rather mysterious group of proteins waiting for a role. The role of the first uncoupling protein to be identified, now known as UCP1, was established in the 1970s in brown adipose tissue (BAT) where it’s physiological role is clearly understood. BAT is found particularly in small mammals at a time in development when non-shivering thermogenesis is vital to maintain core body temperature. Activation of the UCP operates through noradrenaline and cAMP which leads to activation of the protein and actively uncouples the mitochondria of the BAT tissue. This causes generation of heat through the futile cycling of protons and the (futile) consumption of ATP. More recently, a family of UCPs were identified, found widely distributed amongst different tissues and species––even in plants. At the last count these ran to UCP 5. There has been considerable debate and speculation about their possible physiological roles. The most compelling case is made that upregulation of UCP2 and 3 in adipose tissue and muscle respectively may contribute in some way to appetite and weight regulation on the one hand (see Clapham et al., 2000; Lowell and Spiegelman, 2000) and also that UCP expression may help to limit the rate of free radical generation under conditions of oxidative stress. Recently, Echtay et al. (2002a,b) found that an increase in superoxide caused activation of UCP2, distinct from the upregulation of UCP2 expression that probably also follows oxidative stress. These observations strongly suggested that increased superoxide, perhaps primarily generated by mitochondria themselves––and the evidence suggested that the superoxide appears on the matrix face of the inner mitochondrial membrane (Echtay et al., 2002b)––activate the UCP, increase the mitochondrial membrane permeability and so reduce the potential, so decreasing the rate of superoxide generation. A recent paper has suggested that the upregulation of UCP2 may play a neuroprotective role in the brain, (Mattiasson et al., 2003). In mice over expressing the protein, infarct size was reduced following a model of stroke. The suggestion is that UCP2 may cause a mild mitochondrial depolarisation that could limit mitochondrial calcium accumulation, reduce free radical generation and so protect cells from injury. The effect can be mimicked by low doses of uncoupler––a phenomenon now described in several models (Stout et al., 1998; Rodrigo et al., 2002). This is a growing an expanding field, and the implication of UCPs in disease remains speculative but seems an area likely to become important as we begin to understand the biology of these fascinating proteins.

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

401

7. Mitochondria as sources and targets of NO The interplay of mitochondrial respiration with signalling by nitric oxide is a recent development which may have fascinating ramifications both for physiological and pathological processes. NO is widely used as a signalling molecule throughout the body, carrying out numerous roles but most notably regulating local vascular tone and blood flow. In general, NO will cause local vasodilation, increasing oxygen delivery. However, it turns out that mitochondrial complex IV (cytochrome oxidase, COX) also has a high affinity for NO, which therefore competes with oxygen and inhibits mitochondrial respiration. As the inhibition is competitive, NO is more effective an inhibitor at low oxygen tensions. However, it is important to recognise that the tissue oxygen tensions, at which most of our cells normally operate, are far lower than arterial oxygen tension, and are often measured at 10–20 mmHg. These low oxygen tensions have no consequences for mitochondrial respiration, as the affinity of COX for oxygen is so very high (estimated at <1 mmHg), and so there is a huge safety margin to sustain mitochondrial oxidative phosphorylation. Nevertheless, at these oxygen tensions NO may compete significantly for COX binding, and so may inhibit respiration. One of the crucial questions of course is the concentration of NO actually reached within the tissues. This itself is contentious but there will also of course inevitably also be tissue gradients both of oxygen and of NO. Most interesting here in the context of this review is the possibility that mitochondria may themselves be generators of NO. Several labs have recently suggested that mitochondria express a specific NOS, and so may represent a source of NO themselves. This would mean that the respiratory chain could be exposed to high local concentrations of NO sufficient to modulate respiration. The question of course is what is the normal physiological role of this enzyme. Does it play a normal role in regulating the rate of mitochondrial respiration? And if so, how is it regulated, what is the value of suppressing respiratory rate and what evolutionary advantage is there in placing the enzyme within the mitochondria while they will equally be exposed to NO generated elsewhere in the cell. One answer to the latter would be the regulation of function by some factor specific to the intramitochondrial environment––say high calcium when mitochondria take up calcium. Ghafourifar and Richter (1997) in one of the first descriptions of a mtNOS, suggested that NO generation from a preparation of isolated mitochondria was stimulated by loading with calcium and caused some fall in mitochondrial potential. Giulivi, on the other hand has suggested that the enzyme apparently has such a high Ca2þ affinity that it is likely to be already upregulated by resting levels of intramitochondrial calcium (Giulivi, 2003). In experiments using isolated mitochondria, there is always some question mark over the purity of preparation, possible contamination by other membrane or soluble fractions. A very recent paper by Dedkova et al. (2004) apparently provides direct evidence for calcium regulated generation of NO from mitochondria by directly imaging NO generation from permeabilised cells. In this study, the NO sensitive indicator DAF-2 was loaded into cells and the cytosolic signal removed by membrane permeabilisation, leaving the dye in the mitochondrial fraction. Modest addition of calcium increased NO production and

402

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

this response was clearly dependent on mitochondrial Ca2þ uptake. Such an observation gives a rationale for the sequestration of the enzyme into mitochondria, as the enzyme will be selectively sensitive to the calcium concentration only within that compartment. The functional consequences of such a pathway remain to be established, but it seems clear that this is an area of expanding interest. The consequences of NOS activity for mitochondrial function remain controversial, largely dependent on the dose of NO that is delivered. It seems that even modest levels of NO that only partially inhibit respiration may still increase mitochondrial ROS production and cause changes in mitochondrial respiration which are irreversible, at least in the short term. This can cause some mitochondrial depolarisation, which, though difficult to quantify accurately, seems modest but sufficient to limit mitochondrial calcium uptake. We (Rakhit et al., 2001) found that pre-treatment of cardiac myocytes with low concentrations of a NO donor caused long lasting modest mitochondrial depolarisation associated with some protection from ischaemia-reperfusion injury. This was associated with a significant reduction in the capacity of mitochondria to accumulate calcium, which, given the importance of calcium accumulation as a precipitant of cell death (see below), may be sufficient to explain the increased resistance of the cells to injury. A further fascinating role for NO and mitochondria has emerged recently. A study by Nisoli et al. (2003) showed that NO generation can stimulate the upregulation of mitochondrial biogenesis. This operates through the cGMP dependent increase in the expression of PGC1a (see above). As discussed above, PGC1a increases the expression of mtTFA and NRF-1, increasing the biosynthesis of mitochondria. It is not yet clear how widely distributed this mechanism might be amongst different tissues, but it is clearly expressed in adipocytes and hepatocytes. Interestingly, Momken et al. (2002) recently reported that mice deficient in eNOS have deficiencies of mitochondrial enzymes, positing a physiological role for this pathway in regulating mitochondrial mass in muscle (see above). The functional effects of modest levels of NO seem to generate a coordinated series of responses which enhance mitochondrial function––an increase in local blood supply bringing oxygen and substrate and also increasing mitochondrial mass. However, when NO is generated in excess, under some pathological stimulus, these same properties may cause tissue damage. High concentrations of NO in the presence of relatively low tissue oxygen tensions may cause inhibition of respiration. Several groups have suggested that, as respiration is inhibited in the presence of oxygen, this will increase the rate of generation of superoxide by the respiratory chain. Superoxide will in turn interact with NO to generate peroxynitrite which is a highly reactive radical species. Peroxynitrite may in turn react with Complex I to cause nitration of thiols, such that the reversible inhibition of Complex IV by NO is replaced by irreversible inhibition of respiration at Complex I. Sepsis, as a pathological entity, is characterised by excessive generation of NO and peripheral vasodilatation with a collapse in blood pressure. Most textbook descriptions of sepsis suggest that tissue damage results from effective hypoxia which is a consequence of reduced perfusion pressure. However, measurements of tissue oxygen tensions show that deep within the kidney for example, local oxygen tension

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

403

is high, not low, as expected if oxygen extraction were normal in the face of reduced perfusion (see for example, Brealey and Singer, 2003). Such measurements suggest that even if tissues are adequately supplied with oxygen, they are unable to use it, as mitochondrial respiration is impaired. If such inhibition is primarily mediated by NO and competition with oxygen, it is reversible and therefore remediable (see Borutaite et al., 2001). Raising oxygen tension and perfusion pressure should overcome the NO mediated inhibition, although there is some debate whether giving extra oxygen might be harmful as it might also increase the rate of radical generation. Once peroxynitrite has caused nitration of thiols, changes in respiratory function are likely to be irreversible, and it is suggested that such a process might underlie the clinical phenomenon of multiorgan system failure, where patients seem to have past a point of no return. Evidence for such a scheme is still incomplete, but measurements of Complex I activity from patients correlate well with clinical state and seem to be a good predictor of outcome. Thus, paradoxically, perhaps, it seems that the clinical signs of sepsis are generated by the toxicity of the body’s intrinsic defence mechanisms and once again the mitochondrion seems to lie at the heart of this issue.

8. Mitochondria and oxygen sensing, tissue oxygen tensions Oxygen delivery to mitochondria is central to health, and is regulated at many points in the complex feedback loops that control respiration. Several major oxygen sensitive processes help maintain the central supply of oxygen––control of respiratory rate, control of gaseous exchange in the lung and regulation of oxygen carrying capacity in the blood. In the regulation of respiratory rate, the partial pressure of arterial Oxygen is sensed by the Type I cells of the carotid body which responds to a fall in arterial oxygen tension with an increased neural output, so initiating compensatory respiratory reflex responses. If you go up a mountain or for some other reason have a low arterial oxygen tension, you feel breathless and hyperventilate. Gaseous exchange is regulated in part through the matching of ventilation with perfusion––there is no point perfusing areas of the lung which receive no air. This is largely achieved by the vasoconstriction of pulmonary vessels in response to a fall in oxygen tension––hypoxic pulmonary vasoconstriction. This process again is a direct and intrinsic responsiveness of pulmonary vessels to hypoxia. Finally, hypoxia switches on a genetic program to manufacture increased levels of haemoglobin and so increase the oxygen carrying capacity of the circulation. The last of these processes is quite well understood and has, as far as I can see, nothing to do with mitochondria. However, both oxygen transduction in the carotid body and hypoxic pulmonary vasoconstriction have been associated with mitochondrial respiration. This notion goes back a long way and is based primarily on the observation that all agents which interfere with mitochondrial respiration are also potent stimuli of the carotid body and of pulmonary vasoconstriction. There also always seemed a certain logic that the major oxygen utilising system––the mitochondria––should be involved in sensing oxygen availability. The main issue has

404

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

always been one of sensitivity––cytochrome oxidase has a very high affinity for oxygen––a massive safety margin so that normal cells and their mitochondria can operate very effectively even when the tissue oxygen tension falls very low. Normal oxygen tension in most tissues is probably of the order of 10–20 mmHg, far from the atmospheric pressure at which we conduct most of our experiments. The carotid body shows a significant response at an oxygen tension of about 20–30 mmHg. It seemed inconceivable that mitochondrial function would be significantly altered at such relatively high oxygen concentrations. People working in this field have struggled over the years to make sense of these observations. Are the pharmacological effects purely coincidental, signs of some odd response to poisoning, or do they signify a true role for mitochondria in the transduction pathway? Perhaps mitochondria should mind their own business and get on with making ATP. But if the mitochondria are not involved, what is the oxygen sensor? Over the years, candidates for this role have auditioned, some have even been given the part for a while, but most seem to have flopped, while the mitochondria seem to quietly plod on regardless. In the early 1970s, Mills and J€ obsis showed data to suggest that the carotid body shows redox changes in a specialised cytochrome aa3 at relatively high oxygen tensions, strongly suggesting that the mitochondria in the tissue must be specialised in some way. This was largely disregarded by the field with suggestions that the real PO2 sensed by the mitochondria in situ in the tissue was probably much lower than that measured in the perfusate (Mills and Jobsis, 1970, 1972). Later suggestions came of a specific oxygen sensitive potassium channel that was completely independent of any metabolic involvement as it was preserved in patches of membrane (Ganfornina and Lopez-Barneo, 1991) and also of an NADH oxidase that might be expressed in carotid body but seems rather to have been due to a high macrophage content in the structure (macrophages have a high level of expression of NADPH oxidase, and chemoreceptor activity is unchanged in knockout mice lacking the NADPH oxidase). Tim Biscoe and I found that in isolated Type I cells of the rabbit carotid body, mitochondrial potential showed graded and significant depolarisation in response to modest changes in PO2 , and other indices of mitochondrial function––changes in NADH and flavoprotein autofluorescence showed changes in parallel, consistent with a slowing of mitochondrial respiration in response to modest hypoxia (Fig. 7 and see Duchen and Biscoe, 1992a,b). In all other cell types that I have ever studied, including sensory neurons, chromaffin cells, cardiomyocytes, central neurons and astrocytes, nothing happens to these variables until the PO2 is close to zero––certainly not at a PO2 of 10–30 mmHg. Very recently, other groups have found evidence for a novel cytochrome a592 in carotid body tissue that is not found in other excitable tissues and that changes redox state at modest changes in oxygen tension (Streller et al., 2002). The issue is not yet resolved, as Ortega-Saenz et al. (2003) have recently suggested that sensing might involve another, non-mitochondrial respiratory chain that shows selective sensitivity to rotenone compared to other inhibitors of mitochondrial respiration (Ortega-Saenz et al., 2003). What also remains a mystery is how that sensing process might initiate the rest of the pathway that involves the suppression of a K conductance (also inhibited by agents that inhibit mitochondrial

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

405

Fig. 7. Mitochondria as oxygen sensors? Measurements of mitochondrial potential (A) and of NADH autofluorescence (B) from the oxygen sensing isolated Type I cells of the rabbit carotid body. The cells were superfused with salines that had been equilibrated with Nitrogen to achieve varying levels of oxygen and responses to successive applications of hypoxic salines have been superimposed and the levels of PO2 achieved are indicated next to each trace. NAD(P)H autofluorescence was excited at 350 nm and measured between 410–490 nm and mitochondrial potential was measured using the dequench of rhodamine 123 as described above. What is most strange about these recordings is that we have never seen any response to any change in PO2 except close to zero in any other cell type. These responses appear quite unique to the carotoid body and are consistent with an effectively reduced affinity of cytochrome oxidase for oxygen. Adapted from Duchen and Biscoe (1992a) with permission from the Journal of Physiology.

respiration––Buckler and Vaughan-Jones, 1998; Williams and Buckler, 2000) depolarisation of the Type I cells, an influx of calcium and transmitter release onto adjacent nerve terminals. The process of hypoxic vasoconstriction has been equally difficult to resolve, and again mitochondrial involvement has been hotly debated. It seems clear that a key part of the hypoxic vasoconstriction involves the closure of Kþ channels in response to hypoxia causing membrane depolarisation, a rise in intracellular calcium and contraction of the smooth muscle. The parallel with the carotid body response is striking, where again Kþ channel closure and a rise in calcium seem seminal in the transduction pathway. In each case the key question seems to be how is the oxygen tension sensed and how is this then transduced to an alteration in opening

406

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

probability of a Kþ channel or a rise in calcium. Once again, the spotlight has fallen on the mitochondrial respiratory chain as an oxygen sensor. In this field, the emphasis has fallen very much on the idea that the major message that links changes in the rate of mitochondrial respiration with changes in intracellular calcium concentration is a change in the rate of mitochondrial free radical generation. Precise mechanism of mitochondrial ROS generation in response to hypoxia remains obscure. In general, ROS generation is proportional to PO2 ––more ROS are generated at a high oxygen tension. However, it seems that according to many measurements, ROS generation in small vessels increases with hypoxia. There are huge problems in making unequivocal measurements of ROS generation and methodological differences and differences in preparations used may account for some of the differences in findings from different labs. However, in one very interesting paper, Michelakis et al. (2002) compared the properties of mitochondria from vessels that constrict (pulmonary) and relax (renal) in response to hypoxia (see above also, section dealing with ROS generation). Just as hypoxia has opposite effects in these vessels, so did inhibitors of mitochondrial electron transport. However, hypoxia apparently caused a fall in ROS generation in mitochondria from pulmonary vessels but an increase in those from renal vessels. I don’t believe that anyone has any real idea how this might work or what the difference might be between the two populations of mitochondria. Nor is it clear how ROS generation modifies K channel activity––whether it is a direct effect or an indirect modulation through changes in GSH/GSSG ratio (amongst others). The issue remains fraught and full of contradictory data and I would refer you to (Ward and Aaronson, 1999; Archer and Michelakis, 2002; Ward, 2003) for some recent reviews.

9. Mitochondria and glucose sensing Really for the sake of completeness, I wish to include a short mention of the role of mitochondria in glucose sensing. Mitochondria play a central role in glucose sensing in the pancreatic beta cell culminating in the glucose dependent secretion of insulin. The beta cell plasma membrane potential is determined predominantly by the opening of a potassium channel which is closed by intracellular ATP (the KATP channel). The simplest scenario in the beta cell is that a rise in glucose increases the rate of mitochondrial ATP generation, [ATP]/[ADP] rises, causes closure of the K channels, the plasma membrane depolarises, opening voltage gated calcium channels. The rise in [Ca2þ ]c causes insulin secretion. The efficacy of TMPD/Ascorbate in raising calcium and stimulating secretion emphasises the dominant role of mitochondrial ATP generation in the pathway, as these supply electrons directly to Complex IV without engaging any of the rest of intermediary metabolism (see Duchen et al., 1993b). Mitochondrial responsiveness in these cells to subtle changes in [glucose] is largely dependent on the specialisation of the whole pathway if glycolysis and the expression of a glucose transporter (GLUT 2) in the cell membrane that does not saturate as plasma glucose rises. The story is not quite so simple, as the KATP channel is also regulated by other metabolites. It is hard to believe that ATP

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

407

really swings as widely as would be necessary to regulate the channel as predicted simply from single channel properties. This topic has been extensively reviewed and this is not the place. However, it remains clear that it is the mitochondrial response to glucose and increased ATP generation that regulates the electrical activity of the plasma membrane so that mitochondrial function appears pivotal in regulating insulin secretion. A secondary component of the mitochondrial contribution appears as a result of the change in calcium. Thus, with the provision of glucose, the mitochondrial supply of NADH through the TCA cycle increases (measurable in single cells as an increase in NADH autofluorescence––see Duchen et al., 1993a,b) and this promotes an increase in mitochondrial membrane potential, increasing respiratory rate and the rate of ATP generation. Thus the rise in [Ca2þ ]c that follows occurs at a time when the mitochondrial potential is increased. This in turn is expected to increase the efficacy of mitochondrial calcium uptake. It has been suggested that the increase in intramitochondrial calcium at this point stimulates the TCA cycle further, as expected, and that this might mediate a further stimulation of ATP generation (Ainscow and Rutter, 2001). In fact, Maechler and Wollheim (2001) have argued that the secondary stimulation is due not simply to an increased ATP production, but to the generation of intracellular glutamate by the increased TCA cycle activity, which in turn somehow stimulates secretion, although the mechanism of this potentiation remains uncertain.

10. Mitochondria and cell death––necrosis, apoptosis and necroptosis? Given the fundamental role of mitochondria as ATP generators and the requirement of cells for ATP to combat the forces of entropy, it seems self-evident that severe damage to mitochondria will cause cell death. Once ATP levels fall, energy dependent processes that include the maintenance of ion gradients, the contraction of muscle, secretion of transmitters, and the maintenance of normal regulation of calcium required for the coordination of calcium signals will inevitably become disordered. Once ionic gradients dissipate, intracellular osmolarity cannot be maintained, and cells will swell and die by a process of necrosis. Some cell types, specially immortalised cell lines, survive reasonably well simply on supplies of ATP from glycolysis and seem not to require their mitochondria for oxidative phosphorylation. HeLa cells are a prime example, and will grow quite reasonably despite complete inhibition of mitochondrial respiration. In the brain, astrocytes seem able to maintain function adequately on glycolytic ATP, while neurons are absolutely dependent on mitochondrial oxidative phosphorylation for normal function, so that we lose consciousness within seconds of oxygen deprivation. Even if this outcome of mitochondrial energy failure is self-evident, this does not diminish the importance of understanding the pathways that lead to such mitochondrial damage, as necrotic cell death plays a profoundly important role in many disease states. One of the key issues here must surely be a proper understanding of the main consumers of ATP. This issue does not seem to have been fully resolved––after

408

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

all, the rate of consumption of ATP and the rate of ATP depletion must be a primary determinant of the time available before an ischaemic episode leads to irreversible injury. It is often stated that one has 3 min to resuscitate a patient with a cardiac arrest before there is significant risk of irreversible brain injury. Presumably that 3 min interval (or however long it really is!) is a reflection of the rate of ATP consumption. It is often assumed that the major ATP consumers must be the plasmalemmal ATPases––the Ca2þ /H+ ATPase or the Naþ /Kþ ATPase. However, major changes in intracellular calcium and/or sodium concentrations should not occur until ATP levels have fallen to critical levels, as the affinity of the ATPases for ATP is very high. In the brain, increased Naþ and Ca2þ flux do occur as extracellular glutamate starts to accumulate, but even that is dependent on low ATP levels, high intracellular Naþ and low Kþ that drives the glutamate transporter in reverse (e.g. Szatkowski and Attwell, 1994). 10.1. Mitochondria as ATP consumers It seems that the mitochondria themselves may be a major ATP consumer in the face of an anoxic or ischaemic environment. It is useful to think of the mitochondrial ATP synthase essentially as a proton translocating ATPase that normally runs as a synthase because it is driven in ‘reverse’ by the mitochondrial proton gradient established by respiration. Thus, once respiration stops, proton pumping by the respiratory chain stops and the mitochondrial potential will begin to dissipate. The rate of dissipation seems to vary between cells, presumably reflecting differences in proton leak (and perhaps differences in the expression of UCPs––see above) and proton flux through the proton channel of the ATP synthase. As this process continues, the electrochemical potential that drives the ATP synthase––a complex function that depends both on the ATP/ADP ratio and the potential––will reach a state where the thermodynamics favour reversal. The enzyme then acts as an ATPase, consuming ATP and pumping protons out across the mitochondrial inner membrane. This serves to slow the rate of decline of the mitochondrial potential. The operation of this sequence can be experimentally demonstrated in several ways (e.g. see Fig. 8). Thus, if we measure mitochondrial potential and make cells anoxic, we see an initial rapid depolarisation which then slows, in some cases to be maintained at a fairly constant level––again this varies between cell types––until the cell becomes ATP depleted, when the mitochondria will depolarise completely. This is easy to see in cardiomyocytes, as ATP depletion is indicated directly by the shortening of rigor contraction. In the presence of oligomycin, to block the ATPase, the mitochondria depolarise rapidly in response to anoxia and the rate of ATP depletion can be dramatically slowed (see Duchen et al., 1993a; Leyssens et al., 1996; Rouslin, 1991). In Fig. 9, I have illustrated this point using the measurement of intracellular magnesium concentration to indicate the rate of ATP consumption. It remains difficult to measure [ATP] directly at the level of the single cell––or even with high time resolution in a population of cells. The latter can be achieved by transfecting cells with luciferase, but this is often difficult using primary cells and impossible if the cells are acutely isolated (as in the case of the cardiomyocytes that we have used). It

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

409

Fig. 8. Effect of the mitochondrial ATP synthase in slowing mitochondrial depolarisation in response to anoxia. These records show changes in mitochondrial potential recorded from Type I cells of the carotid body but similar records have been obtained from other cell types (chromaffin cells, sensory neurons). Rhodamine 123 was used in the dequench mode and responses to anoxia followed using microfluoriometry. Note that the depolarisation and repolarisation (on reoxygenation) phases are fitted well by the sum of two exponential processes (A, B) with time constants of 2.7 and 55.5 s (depolarisation) and of 1.98 and 99.56 s for repolarisation. In the presence of oligomycin (2.5 lg/ml), the slow components were lost, and the rapid depolarisation and repolarisation were now well fitted by single exponentials corresponding to the fast components (in this example, 23.4 s onset and 2.24 s offset) of the ‘control’ responses. The fast component must represent a leak conductance, while the slower component represents the activity of the F1-F0-ATP synthase running in ‘reverse’ mode effectively opposing the depolarisation. On reoxygenation, the respiratory activity quickly restores potential but the accumulated ADP drives the ATP synthase causing a slower more sustained depolarisation until it is converted back to ATP. This slower phase is also abolished by oligomycin. Note the huge difference in amplitude of the response to anoxia with and without oligomycin, emphasised by superimposing the control trace (dotted line) on that with oligomycin. Adapted with permission from J. Physiol. (Duchen and Biscoe, 1992a).

turns out that the bulk of intracellular Mg2þ is bound to ATP and that ADP has a much lower affinity for Mg2þ than ATP. Thus, as ATP is hydrolysed, [Mg2þ ]c rises (see Leyssens et al., 1996). In the example illustrated, [Mg2þ ]c was measured in cardiomyocytes simultaneously with NADH autofluorescence (as an internal control). On application of FCCP, we see a rapid increase in [Mg2þ ]c simultaneously with the predicted fall in NADH autofluorescence. The change in [Mg2þ ]c was blocked by oligomycin, demonstrating the importance of the F1 F0 -ATPase in determining the rate of ATP consumption under these conditions. In a strange way, although simply a consequence of thermodynamics, this seems to be a manifestation of the evolutionary origin of the mitochondria as independent

410

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

Fig. 9. Mitochondria as ATP consumers. The small inset diagram (A)(i) recapitulates that of Fig. 2 showing the normal coupling of respiration and phosphorylation. When mitochondria depolarise, the ATPase can reverse consuming glycolytic ATP and pumping protons, so maintaining a potential (A)(ii). This is illustrated experimentally in (B), in which [Mg2þ ]c was measured simultaneously with NAD(P)H autofluoresence from single cardiomyocytes. Depolarisation of the mitochondria with FCCP causes an immediate increase in respiration signalled by the decrease in autofluorescence (see above). This is accompanied by a rapid increase in [Mg2þ ]c as ATP is hydrolysed (see text). This continues until it reaches a plateau and the cells go into rigor contracture, signalling the depletion of ATP. On washout of the FCCP, the cells relengthen as ATP is resynthesised and the [Mg2þ ]c recovers. The role of the ATPase in promoting ATP depletion is demonstrated in ii, in which the experiment was repeated in the presence of oligomycin. Adapted with permission from Journal of Physiology (Leyssens et al., 1996).

structures. The mitochondria consume ATP ‘in order’ to maintain their potential, even though in the process they cause damage to their host cells. Interestingly, nature seems to have evolved a defence mechanism to protect the cell from these rogue mitochondria. A protein called IF1, described in the early 1980s, acts to inhibit the reverse mode of the ATPase. IF1 binds to the enzyme as it switches from ATPsynthase to ATPase and inhibits its action (Cabezon et al., 2003). Thus the cell prevents the mitochondria from consuming the reserves of ATP that are needed by the cell, while

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

411

allowing the mitochondrial potential to dissipate. This is a ‘catch 22’ situation for the cell, as mitochondrial depolarisation brings with it the risks of mitochondrial swelling and release of Cytochrome c––see below. As yet, remarkably little is known of the functional consequences of IF1 action, of its expression in different tissues or different species or its impact on cellular pathophysiology. Studies by Rouslin and his group, in the 1980s showed that IF1 function was demonstrable in the hearts of larger mammals with relatively slow beating hearts (rabbit, dog cat) but not in small rodents with fast beating hearts (mouse, rat; see Rouslin, 1991 for review). Thus in cells expressing IF1, the rate of ATP consumption by mitochondria during periods of anoxia was significantly slowed compared to cells in which IF1 was not expressed. The ramifications of the expression of this enzyme in defining for the response to anoxia are quite complex, but interesting to consider, and deserve further exploration. 10.2. The mitochondrial permeability transition pore (mPTP) The mPTP is a large conductance pore formed apparently through a conformational change of several constituent proteins of mitochondrial membrane. The pore opens most clearly under specific and usually pathological conditions, and spans the inner and outer mitochondrial membranes. The exact structure of the mPTP has proven elusive, although elements of the structure seem to include the adenine nucleotide translocator (ANT) in the inner mitochondrial membrane, the voltage dependent anion conductance (VDAC) in the outer membrane, and cyclophilin D, in the matrix, which confers sensitivity of the complex to cyclosporin A, which closes the pore (see Crompton, 2000). Other proteins––the antiapoptotic Bcl-2 and the ‘peripheral type’ benzodiazepine receptor––may be associated with the pore. The channel (once called the ‘mitochondrial megachannel’) is characteristically opened by a combination of high [Ca2þ ]m , oxidative stress, ATP depletion, high inorganic phosphate (Pi) and mitochondrial depolarization (for a recent review, see Crompton et al., 1999). The conductance is inhibited by Mg2þ , ATP, ADP, cyclosporin A (CsA) and a relatively recently discovered agent, sanglifehrin (Clarke et al., 2002) and by modulators of the adenine nucleotide translocase (e.g. bongkrekic acid). It is currently unknown whether the mPTP plays any normal physiological role in mitochondrial homeostasis or in cell physiology, or whether it represents only a pathological process––there is some evidence in support of both proposals (see Kowaltowski et al., 2000; Crompton, 1999). Patch clamp recordings from mitochondrial membranes or work with lipid bilayers has strongly suggested that the channel can exist in multiple conformations, including a low conductance mode (Zoratti and Szabo, 1995; Kinnally et al., 1996; Loupatatzis et al., 2002); and it seems possible that the consequence of its opening may be defined by the conductance states adopted (Ichas and Mazat, 1998). Thus, transient reversible opening of a low conductance state may have quite different consequences to irreversible opening of a high conductance pathway, which inevitably will cause mitochondrial swelling, cytochome c release, caspase activation and apoptotic cell death (see De Giorgi et al., 2002), or the collapse of the mitochondrial potential, ATP consumption and depletion and energetic collapse followed by necrotic cell death.

412

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

One of the biggest problems in studying the mPTP is trying to find a tool that will allow unambiguous experimental verification that the mPTP is involved. CsA is used routinely, binding to Cyp D and inhibiting pore opening. But CsA binds with all cyclophilins and at least 9 such proteins have been identified, although their functions remain uncertain. For example, the cytosolic cyclophilin (Cyp A) binds the ubiquitous signalling protein, calcineurin. Methyl-valine cyclosporin (mv-Cs) is supposed to be more selective as the mv-Cs:cyclophilin complex does not interact with calcineurin, but in truth, we do not know what else mv-Cs might do––it is not as easy an answer as it might seem and indeed, it is not readily available anyway. Another agent, sanglifehrin, has been identified recently which appears to interact with Cyp D at a distinct site from that occupied by CsA and also does not bind calcineurin (Clarke et al., 2002). In an elegant assay, Nieminen et al. (1995) and Petronilli et al. (1998) have each devised techniques to image mPTP opening based on the redistribution of the fluorescent molecule, calcein, between mitochondrial and cytosolic compartments as a way to identify opening of a large conductance pathway. The argument basically suggests that as a large molecular weight compound, calcein should not be able to pass freely across the mitochondrial membrane. Introduced into the cell as the membrane permeant acetoxy-methyl ester, the AM ester is cleaved in the cytosol and/or within the mitochondria. The mitochondrial entrapped dye should not be able to leave, while cytosolic dye should not be able to enter mitochondria. Lemasters’ group have followed the movement of calcein from the cytosol into mitochondria, seen as negatively stained structures that appear dark against the bright cytosolic calcein of hepatocytes. This only works in cells with nice fat mitochondria that don’t themselves accumulate calcein-AM, as simple optical constraints limit the ability to view fine mitochondrial structures in most cells by negative staining (Lemasters and Nieminen, 1999). Petronilli et al. (1998) allow calcein AM to permeate throughout the cell, and find that mitochondria also take up and cleave the AM ester so that they are stained. They simultaneously quench the cytosolic dye with cobalt, leaving the mitochondria brightly stained (see Fig. 10A) so that mPTP opening causes calcein efflux (or cobalt influx?) with a loss of the mitochondrial signal. Anecdotally, we know of a number of labs, including our own, who have attempted these approaches in other cell types in which mitochondria have simply refused to load with calcein. Thus, imaging the pore opening in living cells may be possible, but modulating pore function with clean, selective, cell permeant pharmacological agents is less straightforward. The basic consequences of mPTP opening must be some loss of mitochondrial potential––a complete collapse of potential if opening is complete and irreversible–– and the efflux of mitochondrial calcium (e.g. see Chernyak, 1997; Brustovetsky and Dubinsky, 2000). It is not at all clear whether the pore plays any other role in the life of the mitochondrion, or indeed whether it even exists except under extremely pathophysiological conditions. Nevertheless, there is mounting evidence to suggest that mPTP opening to some degree is a normal feature of mitochondrial physiology. Kowaltkowski et al. (2000) found that mitochondrial potential can be enhanced by manoeuvres that inhibit spontaneous mPTP opening, suggesting that the mPTP may flicker open and closed under resting conditions. Mitochondrial integrity has been

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

413

Fig. 10. Experimental reflections of mPTP opening. (A) Mitochondrial localisation of calcein can be demonstrated using Cobalt to quench the cytosolic signal, here in a cardiac ventricular myocyte. (B) imaging isolated mitochondria loaded with TMRM (40 nM) on the stage of a confocal microscope, transient loss of signal is seen routinely, often followed by the complete loss of signal. This has been interpreted as spontaneous opening of the mPTP (see Huser et al., 1998). (C) When imaging intact cardiomyocytes loaded with TMRM (here at dequenching concentrations), spontaneous transient signals were also seen culminating eventually in global loss of mitochondrial potential. This is seen here as a wave of depolarisation that progresses slowly across the cell, and is followed by a rigor contracture as the mitochondrial ATPase consumes all remaining ATP. Note the loss of mitochondrial integrity at 238’ and the increasing signal at one end of the cell. At 338 s, all mtiochondria are depolarised and by 763 s, the cell has gone into rigor (adapted from Hausenloy et al. (2003)) (D) patch clamp recordings from intact mitochondria within nerve terminals may reveal large conductance channel activity that is activity and calcium dependent and which is modulated by CsA, possibly reflecting in situ activation of the mPTP (see Jonas et al., 1999; trace kindly provided by Elizabeth Jonas).

monitored using the fluorescence indicator calcein. Both approaches discussed above have been used. Using cobalt to quench cytosolic calcein fluorescence, cells have been loaded with calcein and then simply left alone. Some CsA sensitive progressive

414

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

loss of calcein fluorescence was seen, again consistent with the spontaneous opening of the PTP––sufficient to cause dye efflux––under resting conditions (Petronilli et al., 2001). These authors argued that spontaneous transient openings of the mPTP are likely to be innocuous while the full opening causes cell death. We have found that imaging mitochondrial membrane potential (using TMRM or related indicators) in almost any cell type that we have studied, we see small apparently spontaneous and reversible changes in mitochondrial membrane potential (Duchen et al., 1998; Jacobson and Duchen, 2002). Such changes have been reported by a number of groups, but there seems to be little agreement on what they really represent, some authors accepting these transients clearly as transient openings of the mPTP (Huser and Blatter, 1999; De Giorgi et al., 2002), while others have found no association with either calcium uptake or mPTP (Buckman and Reynolds, 2001; Vergun et al., 2003; O’Reilly et al., 2004). Similar transient changes in mitochondrial potential can be seen in isolated mitochondria loaded with a potential sensitive indicator and studied using confocal imaging (see Fig. 10B and Huser et al., 1998), but again the interpretation of these events, initially accepted as opening of the mPTP, has been recently questioned (Vergun et al., 2003). We have clearly found that the incidence of such events in cells increases with illumination of the dyes used to measure potential, and found that this was prevented using antioxidants, strongly arguing that, whatever the fundamental unit in the ‘resting’ cell, oxidative stress induced by illumination enhances the process (Duchen, 2000; Jacobson and Duchen, 2002). This was also clearly sensitive to a variety of inhibitors of the mPTP, including methyl-valine cyclosporine and sanglifehrin, (Hausenloy et al., 2003), and so we have accepted that the transients seen during illumination represent transient mPTP openings. After a period of illumination, these give way to a profound and widespread loss of mitochondrial potential that leads to cell death (Fig. 10C). We found that minimising illumination intensity and so restricting the responses to transient local depolarisations caused no change in cell fate 24 h later, arguing that transient mPTP opening is innocuous. There are data suggesting that the pore may operate spontaneously in ‘resting cells’ under physiological conditions. The evidence is mixed, as there is always a worry that the very act of making the measurements may interfere with the process being measured and so induce it’s appearance––a physiological variant of Heisenberg’s uncertainty principle. Thus, Jonas et al. (1999) have made some patch clamp ‘ramming’ experiments from synaptic terminals in which a patch clamp pipette enters the cytosol and patches intracellular membranes (Fig. 10D). In such recordings apparently from mitochondrial membranes, single channel activity of a low conductance was seen at rest, with increased higher conductance activity associated with synaptic stimulation. The functional significance of such events remain unclear and it has to be said that major loss of mitochondrial potential is not observed at synaptic terminals in response to stimulation, and neither is the loss of mitochondrial calcium one might expect to see (e.g. see Billups and Forsythe, 2002), although it is not clear how readily calcium permeant the subconductance states might be. It remains unclear to us whether the mPTP really does open spontaneously when the cells are not being studied, what its significance might be, or indeed, whether there might be

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

415

several different processes each causing similar transient changes in mitochondrial membrane potential. Further evidence for the opening of the mPTP as a transient reversible events comes in a paper by Petronilli et al. (1999) who suggest that transient opening of the pore, perhaps in a low conductance state, (although this is impossible to verify experimentally in this model) will cause immediate quenching of mitochondrial calcein, while the mitochondrial membrane potential (measured with the potentiometric indicator TMRM) is more or less maintained. The latter is only feasible if the pore opening is brief and transient, so that transient depolarisations can recover. This process, induced by the calcium ionophore A23187, was not associated with cyt c release. However, another mPTP inducer, arachidonic acid, apparently caused full and irreversible opening of the pore, with the collapse of mitochondrial potential and the loss of calcein signal, followed by cytochrome c release. These data appear to contrast with those of Szalai et al. (1999) who showed that exposure of cells to sublethal concentrations of ceramide followed by a transient increase in [Ca2þ ]c caused a transient mitochondrial depolarisation due to transient opening of the mPTP and cyt c release. They argued that the mPTP acts as a detector of the coincidence of a calcium signal with a pro-apoptotic stressor––oxidative stress, ceramide. Perhaps, in this instance, the pore opened fully but reversibly, while in the experiments of Petronilli et al. (1999) the pore opened in a flickering low conductance state? Indeed, Pastorino et al. (1999) also suggested that low concentrations of the pro-apoptotic protein, bax can induce PTP opening in a low conductance state, release cyt c and initiate apoptosis. It is hard to see how a pore that spans both membranes can release proteins from the intermembrane space without causing mitochondrial swelling (see above). Perhaps one needs to invoke some feedback process whereby bax causes cyt c release, perhaps also pro-caspase-9, activating caspase activity and perhaps then promoting mPTP opening? These questions are very hard to resolve within cells, and seem a little bewildering. In particular, it seems that we badly need more precise molecular or pharmacological tools to manipulate the mPTP rather then the miserably blunt tool provided by CsA. There is strong and compelling evidence for a role of the mPTP in cell death on reperfusion in the brain and the heart. Conditions for pore opening are almost ‘perfect’ at the reperfusion of ischaemic heart and probably other tissues (see below)––[Ca2þ ]m is high, the ATP/ADP ratio depressed and Pi high, rates of free radical generation increased and there is a state of oxidative stress, and so it seems most likely that mPTP opening makes a major contribution to cell death under these precise conditions. As a potential pharmaceutical target, it is of great interest to be clear whether this process contributes significantly to cell death in pathological processes in vivo. There is certainly evidence for this both in the heart (for a recent review, see Halestrap, 1999) and in the brain (Friberg and Wieloch, 2002, and see below). 10.3. Apoptosis The last 8 years have brought with them an extraordinary revolution in our thinking about mitochondrial biology. Mitochondria have transformed from being

416

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

the textbook ‘powerhouse of the cell’, to play a central role in defining cell life and cell death. Just 10 years ago, many took the attitude that chemiosmotic principles were solved, the mechanism of ATP synthesis was understood at the molecular level, what more could there be? And yet now it is clear that mitochondria are intimately involved in the delicate processes that sustain the balance between cell life and cell death. In 1994–1995, a cluster of publications emerged demonstrating that cytochrome c (cyt c) when released from the intermembrane mitochondrial space, can initiate an enzyme cascade that results in organised coordinated cell destruction (for an unusual review, see http://www.ergito.com/gtexpts/wang.htm). In this field, the mitochondrion has moved centre stage and has become the focus of research in many labs worldwide. There are so many recent reviews by experts in this field it would be foolish for me to try to reproduce a full scale review of this topic here. I will focus rather on a few issues that deal specifically with aspects of mitochondrial biology pertinent to recent ideas about mitochondrial roles in apoptosis. Key questions remain, although recent studies have come a long way in answering them: ‘How is cyt c release regulated? How is it initiated? Indeed, how is it prevented? Are there multiple routes and mechanisms that trigger release? Can partial cyt c release from mitochondria in one part of a cell have specific local effects or is it an all or none global phenomenon? As cyt c is central to both oxidative phosphorylation and apoptotic signalling, what determines whether the route to cell death upon cyt c release is apoptotic or necrotic? Are there yet more mitochondrial proteins involved in initiation and coordination of apoptosis?’ Cytochrome c is released from the mitochondrial intermembrane space (MIMS), possibly from a pool distinct from that committed to the respiratory chain (Bernardi and Azzone, 1981). Cyt c binds to Apaf-1 and activates procaspase-9, initiating the cascade of cellular self-destruction. In many cells, however, there are other proteins, the inhibitors of apoptosis (a family of proteins, the IAPs), presumably there to provide the belt and braces of security against accidental caspase activation. So, perhaps it is not sufficient just to release cyt c? There, also in the intermembrane space lies another protein, diablo, or smac (second mitochondria-derived activator of caspase), which inhibits the inhibitor, and so will permit activation of the apoptotic pathway to proceed. It turns out that other procaspases also lie quietly in the intermembrane space waiting their chance to participate in this process––procaspase-9 has been identified in the intermembrane space, as well as the flavoprotein apoptosis inducing factor, AIF. Interestingly, several of the other major mitochondrial proteins involved in apoptosis also seem to have routine actions in mitochondrial function, while having particular roles in apoptotic cell death. Thus, cyt c and AIF are both involved in redox reactions, although a specific role for AIF has yet to be identified. One of the key questions is, ‘how do the proteins escape from the intermembrane space?’ For some years, the focus has been the mPTP (see above). Indeed, this was largely regarded as a biochemical oddity or even an artefact of isolated mitochondria and was largely ignored until the suggestion that it might play a role in apoptosis. The question still seems to be, does it? As discussed above, the pore seems to be a pathological configuration of mitochondrial proteins that span the two mitochondrial membranes––

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

417

VDAC (voltage dependent anion channel) in the outer membrane, and ANT (the adenine nucleotide translocase) in the inner, together forming some kind of complex associated with and regulated by the mitochondrial cyclophilin D. It is important to understand that the open pore spans both membranes––it does not per se represent a route of exit for proteins from the intermembrane space except by inducing mitochondrial swelling and rupture of the outer membrane. In most models (but not all) such swelling and rupture has not been seen. The pore has a large conductance. Opening will inevitably cause a collapse of the mitochondrial membrane potential required to drive oxidative phosphorylation. In many cases of receptor driven apoptotic cell death, the mitochondrial membrane potential seems to be maintained until after cyt c release, and so the consensus seems to be that mPTP opening is unlikely as a primary trigger to cyt c release in coordinated growth factor mediated apoptotic cell death. VDAC and the ANT are necessary for transport of molecules across mitochondrial membrane, but also seem to play a pivotal role in forming the mPTP. One can, however, ask the converse question––if a pathway opens the mPTP–– notably, cellular Ca2þ overload and oxidative stress, features of pathological states such as reperfusion injury in the brain and heart––will that then cause cyt c release and apoptosis? A vast literature argues strongly that it can––innumerable inducers of mPTP opening cause apoptosis (Zamzami et al., 1996). What seems remarkable is that overexpression of bcl-2 could protect cells under these conditions. Does this mean that Bcl-2 associates with the mPTP? Bax seems to interact with VDAC to regulate cyt c release (Shimizu et al., 1999), perhaps through the channel forming dimerisation of this erstwhile bacterial pore former. And VDAC appears after all to be a component of the mPTP, and yet it is also clear that bax mediated cyt c release can proceed without any apparent loss of mitochondrial membrane potential (BossyWetzel et al., 1998). What is also exciting is the recent suggestion that the role of mitochondria in apoptosis may be related in part to the control of mitochondrial morphology discussed above. Thus, many have noted mitochondria become increasingly fragmented during apoptosis. In a recent study, a novel approach to quantifying the fragmenatation of mitochondria labelled with a GFP revealed that apoptosis was accompanied by an increase in fragmentation, attributed to inhibition of fusion, that seemed to coincide with the translocation of bax to the mitochondria independently of caspase activation (Karbowski et al., 2004). This is consistent with previous reports from the same lab who showed that suppression of Drp1, the protein involved in mitochondrial fission, inhibited the progression to apoptosis (Frank et al., 2001). A related very exciting study has also shown recently, in a different model, that an increase in mitochondrial fission may lead to increased apoptosis, linked in this work to neurodegenerative disease. Bossy-Wetzel et al. (2003) found that a mutation in a protein known as OPA-1 (related to a yeast protein Mgm1p, that is required to maintain fusion competent mitochondria (Wong et al., 2000), is mutated in a human disease that causes optic atrophy. These authors suggested that mitochondrial fission––a failure of fusion––might recruit apoptotic mechanisms leading to retinal degeneration. The conceptual switch in our thinking from a focus on the energetic role of mitochondria to the regulation of death is perhaps most dramatically exemplified by

418

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

a study describing mitochondrial function in eosinophils. Mitochondria are the ultimate symbiont––bacterial structures that have become so integral to eukaryotic cell life that their function is essential to cell life, while the mitochondria themselves are no longer capable of independent existence. A study by Peachman et al. (2001) suggests that in some instances, perhaps the mitochondrion has become almost a parasite rather than a symbiont, becoming a store room maintained at some energetic cost to the cell in which to place dangerous proteins. Apparently, eosinophils contain only a few dozen mitochondria compared to the thousand or so found in hepatocytes. The cells show almost no measurable oxygen consumption and energetic demands are met by glycolytic ATP production. Thus, inhibition of cytochrome oxidase with cyanide has no effect on ATP levels––the cells survive in CN– for 24 h without loss of ATP! The mitochondria of these cells therefore seem functionally redundant in terms of bioenergetics. However, they maintain a membrane potential maintained by the mitochondrial consumption of ATP. Thus, inhibition of the mitochondrial F1 F0 -ATP synthase with oligomycin or inhibition of the mitochondrial adenine nucleotide translocase (required for mitochondrial ATP transport) caused the collapse of Dwm and apoptotic cell death. As discussed above, the ATP synthase is a proton pumping ATPase that is normally driven as an ATP synthase by the proton gradient across the mitochondrial inner membrane. In the absence of respiration, it will run as an ATPase, consuming ATP, pumping protons out of the mitochondrial matrix and so generating a potential (see above). Incidentally, the same process should cause ATP depletion once the mPTP has opened, raising the question ‘how is the ATP necessary for apoptosis sustained’ if the mPTP is involved? The implication of these observations is that the mitochondria in these cells represent an energetic cost to the cell, consuming glycolytic ATP to maintain a potential. Why should this mechanism have evolved in these cells where oxygen is freely available? It seems as though respiration is switched off in the eosinophils for some reason, perhaps as a means to give a cell in which apoptosis needs to be readily controlled more control over Dwm ? Nor is it clear why the cells should apoptose when Dwm collapses––perhaps mitochondrial swelling is then sufficient to release cyt c? What is so remarkable is that here the mitochondria seem to have lost any useful function in terms of bioenergetics, and are present solely as a safe house in which to store away proapoptotic proteins, maintained by the cell at an energetic cost. Skulachev (2000) has made the interesting point that the MIMS is evolutionarily a sort of no-man’s land––the inner membrane is thought to have originated from bacterial membranes, the outer membrane as a delimiting membrane from the host cell, and this space is emerging as a fascinating microscopic intracellular compartment with it’s very own microenvironment that is crucially controlled and involved in regulation of cell life and death.

11. Mitochondria and neurodegenerative disease The neurodegenerative diseases represent a major cause of concern in the health professions today. These diseases are profoundly debilitating, and include Alzhei-

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

419

mer’s disease, Parkinson’s disease, Motoneuron disease (amyotrophic lateral sclerosis, or ALS, also known as Lou Gehrig’s disease in the USA), multiple sclerosis (MS), injury to the CNS through chronic low grade hypoxia, the rarer but crippling Huntington’s disease, Wilson’s disease and Freidreich’s Ataxia. A number of inherited disorders of the mitochondrial genome also cause disorders of the CNS, known generically as the mitochondrial encephalomyopathies. There are, of course, more. In all of these diseases, the fundamental pathophysiological mechanisms remain unclear, but in all, disorders of mitochondrial function have been implicated at some level of the pathogenic process. A major challenge of contemporary neuroscience is surely to understand the extent to which these changes in mitochondrial function represent primary or secondary components of the pathophysiological process, and to understand the basic pathways that lead to the development of disease. The causes of neuronal dysfunction and cell death are varied and may follow a number of distinct pathways in each of these disease processes. As discussed above, it is self-evident that mitochondrial dysfunction will manifest itself as cellular dysfunction or death. Impaired ATP generation will cause a failure of cellular homeostasis, with attendant changes in the ionic balance for Naþ , Kþ , Cl– and Ca2þ that will disturb the patterning of electrical signals and of the intracellular [Ca2þ ] ([Ca2þ ]c ) signals that together underpin the transmission of information in the CNS. Ultimately, ATP depletion will lead to necrotic cell death. Other forms of mitochondrial injury may lead to the release of pro-apoptotic factors, particularly of mitochondrial cytochrome c and the initiation of the cascade to apoptotic cell death. The process of glutamate excitotoxicity is well recognised as a cause of neuronal death most notably following periods of anoxia. This has become a prototypical model for cell death induced by calcium overload and oxidative stress, and there is a widespread view that the same processes may underlie more insidious cell death in at least some of the neurodegenerative diseases. I will therefore address what we know of the involvement of mitochondria in acute glutamate neurotoxicity and will also attempt to examine the evidence for such processes in the neurodegenerative diseases. While the importance of these processes in neuronal signalling may be selfevident, our understanding of the function of glial elements and their contribution both to normal signalling in the CNS and to the development of pathology has been growing rapidly in recent years. The contribution of mitochondrial function to glial physiology and of mitochondrial dysfunction to glial pathophysiology is an emerging field of which we still understand relatively little, and the interplay between neurons and glia in defining patterns of cell injury is a topic of increasing interest. I will accept also that this is not necessarily a balanced account of all these disease processes, but rather reflects our own research interests and contributions. For a specific overview of recent studies on the involvement of mitochondrial disorders in neurodegenerative disease, see the excellent review by Nicholls (2002). 11.1. Mitochondrial calcium overload as trigger of neuronal death It has long been clear that elevation of the concentration of free cytosolic Ca2þ ([Ca2þ ]c ) is critical for many types of neuronal cell death (Choi, 1992). The most

420

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

intensively studied model of cell death induced by calcium overload must surely be that of excitotoxic cell death, whereby neurons die in response to prolonged exposure to the excitatory amino acid, glutamate. This is usually studied as a model of the cell death that extends the penumbra of a stroke, as glutamate can accumulate to high concentrations in ischaemic brain spreading to areas outside the focus of acute ischaemia (Szatkowski and Attwell, 1994). It is widely believed that some more subtle form of the same process may contribute to significant pathology in a more insidious chronic process to cause neurodegeneration in other disease models. Early studies assumed that large elevations of [Ca2þ ]c per se were toxic, a phenomenon sometimes referred to as ‘Ca2þ overload’. More recent experiments have suggested that rather than a simple and inevitable consequence of a high global [Ca2þ ]c , the route of calcium entry may play a crucial role in defining the outcome, although this remains slightly contentious. Such ‘source specificity’ seems arise as a consequence of microdomains of high [Ca2þ ]c established in critical regions of the cell where activation of localised enzyme systems may play an important role in triggering neuronal death (see Tymianski et al., 1993; Sattler et al., 1999 and Sattler and Tymianski, 2001 for a review). In many experimental systems, (e.g. in hippocampal neurons and cerebellar granule cells), glutamate neurototoxicity depends primarily on the activation of NMDA receptors, which are the most highly Ca2þ permeant of the subtypes of glutamate receptor. Features of NMDA mediated excitotoxicity include extracellular Naþ dependent cell swelling and a Ca2þ -dependent process leading to cell death. Cell death can be prevented by removal of extracellular Ca2þ , firmly establishing raised [Ca2þ ]c as a trigger for excitotoxic neuron death, and by blockade of NMDA receptors. Following glutamate application [Ca2þ ]c follows a stereotyped pattern. [Ca2þ ]c rises transiently, then staying at a plateau or with an incomplete recovery towards baseline. Following a variable latency, [Ca2þ ]c rises again, probably into the lM range, closely accompanied by the collapse of mitochondrial potential. Once this has happened, [Ca2þ ]c is independent of external Ca2þ , and either the plasmalemmal Ca2þ ATPase is non-functional, or ATP has been depleted, preventing its operation. The presence of glutamate is only required for the initial [Ca2þ ]c transient, although some of the latter phases of the response are also sensitive to NMDA receptor antagonists (e.g. see Vergun et al., 1999) suggesting that glutamate release, either in response to reverberating circuits in the cultures or due to reversed uptake contribute to sustained Ca2þ influx. Nevertheless, it appears that the initial transient in [Ca2þ ]c rise is pivotal in activating a Ca2þ -dependent process responsible for the necrotic cell death. Under conditions where [Ca2þ ]c is elevated due to activation of Ca2þ permeant excitatory amino acid (EAA) receptors, in most cases NMDA receptors, intramitochondrial Ca2þ ([Ca2þ ]m ) can rise substantially and it has even been suggested that mitochondria are somehow selective targets for NMDA induced Ca2þ influx (Peng and Greenamyre, 1998). Recent evidence has suggested that the rise in intramitochondrial [Ca2þ ] plays a central role in initiating the progression to neuronal death under these circumstances. Thus, limiting mitochondrial Ca2þ uptake simply by depolarisation of mitochondrial potential protects neurons from delayed cell death following NMDA exposure (Nicholls and Budd, 2000; Stout et al., 1988). These experiments also show that the collapse of Dwm per se cannot be the primary cause

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

421

for delayed neuronal death following NMDA application, which must be somehow a consequence of the process initiated within mitochondria by the accumulated Ca2þ that causes the collapse of Dwm . In the presence of physiological amounts of phosphate, it is estimated that mitochondria can accumulate and retain mM concentrations of total Ca2þ (Nicholls and Budd, 2000). However, excessive Ca2þ accumulation has at least two deleterious effects, it can lead to the opening of the mPTP, and to the deposition of Ca2þ phosphate precipitates. It has also been suggested that high [Ca2þ ]m may increase the rate of mitochondrial ROS generation (Dugan et al., 1995 and see above) although more recent data suggest that calcium alone may not suffice but may need also to be associated with some damage to the respiratory chain (Votyakova and Reynolds, 2001). 11.2. Inducers of neuronal damage: the role of free radical species It seems clear that while accumulation of mitochondrial calcium alone is required it is not sufficient to cause cell death in these models of cellular calcium overload. There are a number of additional factors that have been invoked as contributors to the extent of NMDA-mediated cell death––the generation of reactive oxygen species (ROS), the status of antioxidant defences in the cell or contributions to the calcium signal through the release of Ca2þ from ER stores. The generation of ROS is closely liked to mitochondrial respiration and mitochondrial function. In addition, nitric oxide (NO) has recently emerged as central factor in defining the source specificity of NMDA mediated neurotoxicity. We have found that inhibition of NOS by L-NAME attenuated the glutamate induced loss of Dwm and NO exposure in combination with an otherwise innocuous Ca2þ load now provoked mitochondrial depolarization (Keelan et al., 1999). Thus, it appears that elevated [Ca2þ ]c acts synergistically with NO to augment mitochondrial depolarization. The ‘source specificity’ for Ca2þ neurotoxicity appears to result from the co-localisation of the NMDA receptors and nNOS through the action of the scaffolding protein PSD-95. Sattler et al. (1999) showed that suppression of PSD-95 expression selectively reduced cell death and NO generation in response to NMDA, without altering the net Ca2þ influx, and they have even remarkably managed to reduce infarct size in intact animals by using cleverly targeted constructs to detach PSD-95 from the NMDA receptor (Aarts et al., 2002). Another interesting speculation involves the possible role of NO generated by an intramitochondrial NOS, although even the existence of such an enzyme remains controversial (see Ghafourifar and Richter (1997) and above). In our hands, in a model of NMDA toxicity in hippocampal neurons, the generation of ROS––superoxide anion, hydroxyl radicals and singlet oxygen––appears to play no part in the collapse of mitochondrial potential (Vergun et al., 2001), although this may not apply to other model systems. Thus, a variety of free radical scavengers had no effect at all on the rate of mitochondrial depolarisation. These studies also revealed that such agents may have a substantial direct effect on the NMDA channels, limiting Ca2þ influx and this alone may account for the neuroprotective actions of these agents. Nevertheless, much previous work has shown that

422

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

ROS are neurotoxic (see Beal, 1998). While we found no evidence of increased ROS production in the short term in hippocampal neurons exposed to glutamate (Vergun et al., 2001) others have shown increases in free radical generation following NMDA application to cultured neurons (Castilho et al., 1999; Lafon-Cazal et al., 1993). In the latter study, using electron paramagnetic resonance, increase in ROS generation was delayed in relation to the time course of mitochondrial depolarisation and so may reflect a separate converging pathway that contributes to cell death. There is a widespread perception that a rise in intramitochondrial [Ca2þ ] causes an increase in ROS generation. The direct evidence for this is actually rather limited, but widely cited. Thus, Dykens (1994) used electron paramagnetic resonance spectroscopy to identify an increase in the rate of hydroxyl radicals generated in isolated mitochondria on addition of calcium. He also showed that the OH exposure caused further complex I damage, suggesting a feedforward cycle of injury. Nevertheless, more recent studies have suggested that mitochondrial ROS responses to calcium may be dependent on the substrate that is being used––mitochondria metabolising succinate show a decrease in ROS generation in response to calcium that is membrane potential dependent. However, if complex I is inhibited, calcium seems instead to increase ROS generation (see Starkov et al., 2002; Votyakova and Reynolds, 2001). The true response of intact mitochondria within the cell to pathophysiological calcium loads remains uncertain, not least because we are uncertain how best to interpret the experimental data. Glutamate is able to cause cell damage and death even in cells that do not express ionotropic (or metabotropic) receptors. This toxic process is mediated by the depletion of glutathione (GSH) by competition between glutamate and cysteine for the cysteine transporter, preventing the supply of cysteine that is required for resynthesis of GSH (Murphy et al., 1989; Schubert and Piasecki, 2001), a process which has been referred to as ‘‘oxidative’’ glutamate toxicity. As GSH functions as a major anti-oxidant, removing ROS, this oxidative toxicity involves an increase in net mitochondrial ROS production, following the depletion of antioxidant defence mechanisms. The extent to which this mode of cell injury contributes to cell death in vivo in the CNS is not at all clear, but seems likely to be modest, given the degree of protection conferred by inhibition of NMDA receptors. Oxidative damage may be particularly evident in cultured neurons, or neurons in vivo, at periods in development when NMDA or other excitatory amino acid receptors are not yet mature. 11.3. Motor neuron disease Motoneuron disease or amyotrophic lateral sclerosis (ALS) is an appalling and debilitating disease which causes the selective degeneration of motoneurons, weakness and death after a few years. We will know remarkably little about the cause of motoneuron degeneration, and understand even less of the mechanisms that make that degeneration selective for one population of cells. There is quiet a widespread view that the primary mechanism of motoneuron injury involves calcium overload and mitochondrial damage.

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

423

While glutamate toxicity is generally associated with a calcium load to the cell through NMDA receptors, sub-populations of AMPA and kainate receptors have high Ca2þ permeability. Such receptors appear to lack the GluR2 subunit of the receptor which blocks calcium permeation. Ca2þ permeant AMPA/kainate receptors are highly expressed in motoneurons and their activation causes mitochondrial depolarization and cell death in motoneurons in culture (see Carriedo et al., 1996, 2000). Furthermore, three toxins (the ‘mussel toxin’, domoic acid; b-L-ODAP; and beta-N-methylamino–alanine (BMAA)) which all cause motoneuron like symptoms in primates and humans all act at non-NMDA, AMPA/kainate receptors. Activation of AMPA and kainate receptors appear to increase the rate of ROS formation in motoneurons but not in nearby interneurons. It has been argued that accumulation of calcium in mitochondria triggers an increase in ROS generation––thus, manipulations that inhibit mitochondrial Ca2þ uptake reduced both ROS generation and motoneuron injury (Carriedo et al., 2000). As different neuronal populations differ in the specific numbers and types of glutamate receptors present on the neuronal surface, it is likely that variations in the extent of Ca2þ influx will occur in response to activation of glutamate-gated receptors. Furthermore, the effect of a given total Ca2þ influx on changes in [Ca2þ ]c will also depend on neuronal calcium buffering power. This varies substantially between populations of neurons, and seems peculiarly low in those populations of motoneurons which are more likely to show degeneration in ALS (see Palecek et al., 1999). It has been suggested that this low buffering power may play a particularly important role in defining the vulnerability of these motoneuron populations in ALS, an argument supported by the observation that neurons can be protected from excitotoxic death or by injury following a peripheral nerve crush by overexpression of the calcium buffering protein, parvalbumin (Van Den Bosch et al., 2002; Dekkers et al., 2004). Thus, motoneurons appear to be vulnerable to excitotoxic injury because they express a high proportion of Ca2þ permeant AMPA/kainite receptors and because they have a low cytosolic calcium buffering power, perhaps increasing the relative importance of mitochondria in Ca2þ uptake and regulation. Nevertheless, it has to be said that this would apply to motoneurons in healthy individuals, and so cannot account alone for the progression of degeneration. A small proportion of patients with ALS have a hereditary familial disease (familial ALS or FALS) and a small proportion of those patients in turn have a mutation of CuZnSOD (SOD1). It has to be stressed that these represent only a tiny fraction of all patients with ALS, and yet it probably represents the only good model of the disease that we have. Such mutations have been replicated in transgenic mice which also develop a selective motoneuron degeneration. When first described, it was assumed that the mutation would impair the function of the enzyme and so the disease would inevitably reflect a failure of cellular antioxidant defence. However, this was not the case. The mutation does not impair the antioxidant capacity of the enzyme, and animals cannot be protected by the overexpression of a normal SOD. Instead, it seems to involve a gain of function, so that expression of the protein is itself somehow toxic to the cells. The mechanism whereby mutant SOD1 causes motoneuron death remains unknown. What is clear is that one of the first signs of disease, before the appearance of symptoms, is a massive degeneration of

424

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

mitochondria (Kong and Xu, 1998), and so mitochondrial damage seems key to understanding the cellular degeneration that follows. What is so strange and remarkable is that the mutant protein is expressed in all tissues, and yet it is only really motoneurons that show significant pathology. Why? Increased emphasis in ALS research seems to be moving towards a focus on the interactions between motoneurons and glia. This view is supported by the remarkable observation that in chimeric mice with cells expressing either normal and mutant SOD1, toxicity to MNs seemed dependent on damage from the SOD1 and mutation expressed in nonneuronal cells (Clement et al., 2003). Thus, nonneuronal (glial) cells that did not express the SOD1 mutation significantly delayed degeneration and extended the survival of MNs that expressed the mutation. Rao et al. (2003) have suggested that a critical step in the neurodegeneration is founded on glutamate toxicity that might be exacerbated by a failure of glutamate uptake by surrounding astrocytes. They found that motoneurons seemed to show greater increases in [Ca2þ ]c and in free radical generation in response to glutamate compared to surrounding spinal interneurons, and also found that impaired astrocytic glutamate uptake was focused around motoneurons. They suggested that that oxidative stress in motoneurons is somehow transmitted to surrounding astrocytes causing protein thiol oxidation and impaired glutamate uptake. This provides the basis for a destructive feedback loop––increased exposure of motorneurons to glutamate increases ROS generation, while the increased ROS generation further impairs glutamate uptake and so increases the exposure of motoneurons to glutamate. We still don’t know here why glutamate should cause ROS generation from the MNs. There has been a widespread view that a mitochondrial calcium load will increase mitochondrial ROS generation. This is based on a small literature (e.g. see Dykens, 1994) and it has to be said that the evidence for such a process in intact cells is really very weak (see above). Rao et al. (2003) have shown quite convincingly that glutamate induced increase in Ca2þ correlates well with an increase in ROS generation and loss of mitochondrial potential, but the causative sequence is less clear––do mitochondria depolarise because of ROS generation? The implication in the work is vice versa––that mitochondrial calcium accumulation causes depolarisation and this increases ROS generation, but, as discussed above, mitochondrial depolarisation tends to decrease electron leak. 11.4. Alzheimer’s disease and the toxicity of amyloid beta protein Alzheimer’s disease is the most common form of dementia, and it is by definition characterised by the accumulation in the brain of extracellular neuritic plaques, together with the presence of intraneuronal neurofibrillary tangles (NFT) and progressive neurodegeneration. The plaques are mainly composed of amyloid-b (Ab), a 39–43 amino acid peptide, forming large insoluble fibrillary aggregates, and are surrounded by dystrophic neurites and activated glial cells. While the role of this peptide in the development of the pathology is not yet clear, excessive accumulation of Ab––either due to excess production or reduced clearance––appears to be sufficient to cause the disease (N€ aslund et al., 2000). Thus, patients with Down’s Syndrome, who have an extra copy of the chromosome carrying the gene for the Ab

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

425

precursor protein (APP), develop plaques and dementia at an early age. Indeed, all the known mutations associated with familial Alzheimer’s disease are in the genes for APP or for the presenilins, enzymes involved in the processing of APP––all of which result in Ab overproduction. The ‘‘amyloid hypothesis’’ has been challenged in view of the fact that in APP transgenic mice, signs of electrophysiological and behavioural changes appear before any significant Ab deposition (Chapman et al., 1999), while the plaque number does not correlate closely with cognitive impairment. In fact, some cognitively intact elderly individuals have been described in whom a considerable plaque load has been found at post mortem. A likely explanation of this apparent conflict is that the pathogenic form of Ab is not the large, relatively unreactive, static plaque. In AD it is the total Ab load, rather than the plaque load, that correlates with the degree of cognitive impairment (Snowdon, 1997). While an impressive number of studies in the last decade have dealt with Ab toxicity in vitro and in vivo, the exact mechanism of toxicity remains elusive. 11.4.1. Ab and [Ca2þ ]c signalling The effects of Ab amyloid on [Ca2þ ]c signalling have been studied by many groups, and it has become widespread practice to talk of ‘calcium dysregulation’ in relation to the disease. Unfortunately, scrutiny of the data show that these studies have yielded a range of conflicting observations without any clear consensus on mechanism or the relationship between alterations in calcium signalling and the pathogenesis of the disease. As calcium signals play such a fundamental role in neurons and in glial cells as integrators of signalling pathways, they represent an obvious candidate either as a mediator or as a manifestation of pathological process in the CNS. Amyloid peptides have been described as increasing [Ca2þ ]c , decreasing [Ca2þ ]c , altering the dynamics of [Ca2þ ]c signals, having no effect at all. From such a confusing literature, it is almost impossible to extract an intelligible picture from which to draw meaningful conclusions. One problem may well arise from the variety of preparations which are used. A number of groups have studied the effects of Ab on cell lines and on non CNS cells such as fibroblasts. Alzheimer’s disease is clearly a disease of the CNS, but it turns out that abnormalities of cell or mitochondrial function can also be identified in peripheral tissues. The significance of these alterations has never been clear. They do not appear to result cause any significant pathology in other tissues, but, if Ab has effects on fibroblasts, this might inform us about some fundamental action of the peptide on cell membranes or aspects of cell function. However, it does not follow clearly that we can extrapolate directly to understand the basis of disease in the CNS. To complicate matters further, it is not clear whether changes in [Ca2þ ]c signalling are a primary or secondary phenomenon––do these changes represent a cause of the disease or are they rather a manifestation of some other pathological process? Given these limitations, I will focus here on a few studies that have explored the effects of amyloid peptides on [Ca2þ ]c signals. What, then, is the evidence that Ab causes alterations in [Ca2þ ]c signalling that might exacerbate, if not directly trigger pathological changes in the disease? [Ca2þ ]c signalling may be affected in many

426

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

ways––all mechanisms involved in cellular [Ca2þ ]c regulation are potentially subject to modulation. Thus, Ab may interfere with calcium influx pathways, primarily represented in neurons by voltage gated calcium channels, ligand gated channels and in astrocytes as capacitative influx channels. It may interfere with the release or storage of calcium from intracellular pools. It may interfere with the clearance mechanisms required to restore [Ca2þ ]c to resting levels after a signal, the Ca2þ -Hþ ATPase or possibly the Naþ /Ca2þ exchanger, either directly or indirectly by altering cellular metabolic state. Observations published to date make all these mechanisms candidates for alteration by Ab. In neuroblastoma cells, Ab increased the amplitude of current through L-type Ca2þ channels and indeed, the authors showed that cells were protected from Ab induced cell death by blockade of L-type channels, suggesting that the toxicity of Ab was mediated by augmentation of Ca2þ influx, and [Ca2þ ]c overload, which triggered oxidative stress and cell injury (Ho et al., 2001). Rovira et al. (2002) also found that Ab increases the Ca2þ current in hippocampal neurons in culture, but while the effect of Ab 25–35 was attributable to activation of L-type channels, Ab 1–40 seemed rather to open other channel types, possibly including N-type channels associated with transmitter release, leading the authors to suggest that these might be neurotoxic by enhancing glutamate release and so contribute to excitotoxic cell death. Green and Peers (2001) also found that Ab can increase the amplitude of L-type calcium currents in PC12 cells but they found that Ab 25–35 and 1–40 seemed to act in the same way. In contrast, Kasparova et al. (2001), showed that chronic exposure of neuroblastoma cells to Ab 1–40 reduced the Ca2þ current carried through N-type channels and suggested that this effect might underlie a failure of neurotransmitter release, altered synaptic activity and that this could account in some measure for loss of cognition. These studies suggested that the primary action of Ab is to alter [Ca2þ ]c signalling. In contrast, in a study of cortical neurons in culture, it was suggested that an Ab induced increase in Ca2þ was attenuated by antioxidants which were also cytoprotective, suggesting that the rise in [Ca2þ ]c is secondary to the oxidant stress induced by Ab (Huang et al., 2000). This group also found that the Ab-induced Ca2þ influx seen in these cells was not attributable to influx through any readily identifiable pathway, suggesting that Ab induces Ca2þ entry through a pore or by some form of alteration in membrane leakiness to Ca2þ (see below). We have found that Ab had no apparent effect at all on [Ca2þ ]c in hippocampal neurons over the course of an hour or so, but that it did cause the appearance of complex [Ca2þ ]c signals in astrocytes attributable to Ca2þ influx through an Ab induced channel (Abramov et al., 2003). In the early 1990’s, it was shown that amyloid b peptides can form channels when incorporated into artificial planar lipid membranes. These channels were essentially cationic with a selectivity Csþ > Liþ > Ca2þ ¼ Kþ and there is substantial data to suggest that these channels may mediate calcium influx into cells in response to Ab. The channels show sporadic activity and seem capable of generating a number of different conductance states (Arispe et al., 1993). The peptide apparently inserts into the membranes of lipid vesicles where it can be visualised through immunofluorescence (Lin et al., 2001) and where it forms a

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

427

calcium permeable channel. The channel is apparently blocked by zinc, although the action of zinc is confusing, as it can also complex with Ab, prevent aggregation and so prevent insertion into the membrane or prevent pore formation rather than blocking the formed channel (see Arispe et al., 1996; Rhee et al., 1998; Bush, 2003). As in all the experiments described here, the channels are formed by Ab 25–35 or 1–40, but not by the reverse peptide, 35–25, confirming that this is specific to the toxic peptides and is not some non-specific general disruption of membrane structure by addition of peptides. Lin et al. (2001) have shown multimeric channel structures in artificial membranes by atomic force microscopy, and the same group have also shown that Ab can induce sporadic transient changes in Ca2þ in endothelial cells which are entirely due to Ca2þ influx. In our own recent study of hippocampal cells in culture, we also found that Ab 25–35, 1–40 but not 35–25 caused sporadic changes in [Ca2þ ]c in astrocytes that were entirely dependent on external Ca2þ (Abramov et al., 2003). Using the quench of the Ca2þ indicator, fura-2, by Mn2þ influx we were able to show that each Ca2þ transient was due to a pulse of Ca2þ influx and that the signals were independent of Ca2þ release from ER stored Ca2þ . Confocal imaging revealed focal points of [Ca2þ ]c elevation in astrocytes followed by the slower spread of the [Ca2þ ]c signal by diffusion. The signals were prevented by zinc, but once established, could not be blocked by zinc. These data provide further evidence to suggest that Ab may form cationic calcium permeable channels in cell membranes and mediate changes in [Ca2þ ]c signaling. What was most surprising about this study was the selectivity of the action of Ab which caused dramatic changes in [Ca2þ ]c signals in astrocytes while having no noticeable effect at all on [Ca2þ ]c signals in adjacent neurons. This is very puzzling. If Ab acts simply by inserting into lipid bilayers and forming a channel, why should there be a difference in different cells? The answer may lie in observations on the importance of lipid composition in the pore forming activity of the peptide. Recent studies have emphasised the importance of the cholesterol content of lipid membranes for Ab channel formation. Thus, the pore forming activity of Ab in bilayers is inversely related to the cholesterol content of the lipid mixture. Similarly, depletion of cholesterol content in cells by treatment with cyclodextrin or inhibition of cholesterol synthesis in PC12 cells increased Ab toxicity (Arispe and Doh, 2002). Similarly, Kawahara and Kuroda (2001) showed that the [Ca2þ ]c increase in cells was attenuated by pretreament of the cells to increase membrane cholesterol content. We have no data at present about the differences in cholesterol content of different cell types in the CNS, but this seems a simple mechanism that might account for differences in the vulnerability of different cell types to Ab toxicity. 11.4.2. Mitochondria and oxidative stress as mediators of amyloid toxicity The fundamental mechansism of Ab toxicity and the biology of Ab are not well understood. However, mitochondrial abnormalities, namely a decrease in mitochondrial mass and reduced mtDNA content, have been identified as a very early pathological sign in AD, preceding the appearance of NFT, specifically in those neurons most vulnerable to degeneration (de la Monte et al., 2000; Hirai et al., 2001). A decrease in cerebral glucose utilisation has been observed in AD, even when

428

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

corrected for changes in volume due to brain atrophy (Ibanez et al., 1998). Recently it has even been found that APP may be mitochondrially targeted in neurons in a transgenic mouse model. The full length molecule interacts with the mitochondrial protein import motors but transport is not complete, causing mitochondrial dysfunction and inhibition of ATP synthesis (Anandatheerthavarada et al., 2003). It might be expected that the limited glycolytic capacity of neurons should make them especially dependent on mitochondrial oxidative phosphorylation. The activity of key mitochondrial enzymes such as a-ketoglutarate dehydrogenase (a-KGDH), pyruvate dehydrogenase (PDH) and COX is decreased in AD (Gibson et al., 1998; Kish et al., 1992; Sheu et al., 1985). In particular the literature showing COX deficiency in AD is abundant, showing reductions in COX mRNA, in protein expression, and decreased COX activity even in peripheral tissues. APP overexpressed in cultured cells leads to decreased COX activity, abnormal mitochondrial morphology and decreased mitochondrial membrane potential (Mutisya et al., 1994). We and others have found that exposure of isolated rat brain mitochondria to Ab causes a decrease in mitochondrial enzyme activity, respiration and membrane potential (Canevari et al., 1999; Casley et al., 2002). It has been shown that Ab induces cytochrome c release and caspase activation only in cells with a functional respiratory chain (Morais Cardoso et al., 2002), suggesting that functional mitochondria play a significant part as targets or mediators of Ab toxicity. Ab potentiated calcium-induced opening of the mPTP (see above) and caused mitochondrial swelling. As discussed above, mitochondria are both targets of and sources of oxidative damage, and dysfunctional mitochondria in particular may be a source of ROS generation. Mitochondrial inhibition, which may increase ROS generation, also increases the amyloidogenic processing of APP, creating the conditions for further cell damage and representing the basis for a destructive cycle of oxidative stress and mitochondrial damage. It has been suggested for many years that the pathogenesis of the neurodegeneration in AD involves oxidative stress in some form. Evidence for oxidative damage to proteins, to DNA and increased levels of lipid peroxidation have all been described in AD brains. Ab has also been shown using a variety of assays to cause increased production of ROS and/or impaired antioxidant defences (by definition, oxidative stress) and ROS mediated damage to cellular structures––peroxidation or nitration––in a number of model systems (e.g. see Smith et al., 1997; Varadarajan et al., 2000; Cecchi et al., 1999). Similarly, considerable evidence suggests that antioxidant strategies may protect cell model systems from Ab induced toxicity (e.g. for review, see Varadarajan et al., 2000). The key questions then are the relative importance of oxidative stress as a contributor to Ab induced neurodegeneration, the mechanism(s) by which Ab causes oxidative stress and the mechanism(s) whereby oxidative stress leads to neurodegeneration. As described above, the activity of a number of glutamine synthetase and creatine kinase, of a-KGDH, PDH and of aconitase (see Kish, 1997; Gibson et al., 2000) is decreased in AD brains and in cells exposed to Ab. These are all enzymes that are very highly sensitive to oxidative modification and that are altered by exposure to a range of pro-oxidants (e.g. see Tretter and Adam-Vizi, 2000). Another system shown

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

429

to be vulnerable to oxidative modification is the glutamate transporter on astrocytes (Harris et al., 1996), and it has been suggested that the failure of the transport pathway might lead to extracellular glutamate accumulation and therefore glutamate excitotoxic neurodegeneration. One major index of oxidative stress is the level of glutathione (GSH; critically, it is the redox state of GSH, or GSH/GSSG ratio that matters), and a loss of GSH has been described repeatedly in a variety of model systems, both in AD brains, in other tissues of patients with familial forms of AD, and in models in which cells have been exposed to Ab (Cecchi et al., 1999; Muller et al., 1997; Abramov et al., 2003). Glutathione (GSH) is one of the major antioxidant systems in the CNS. Astrocytes have a substantially higher GSH content than neurons and appear responsible for the maintenance of GSH levels in neurons by releasing GSH which is broken down by an ectoenzyme into a dipeptide that can then be used by neurons to maintain their own levels of GSH (Keelan et al., 2001; Dringen and Hirrlinger, 2003). We used confocal imaging of [GSH] in mixed hippocampal cultures and found that exposure to Ab for 24 h caused depletion of GSH in both astrocytes and in neurons, identifiable separately in the imaging experiments, even though the changes in calcium signaling described above were confined to the astrocytes. GSH depletion in both cell types was Ca2þ dependent, suggesting that the alterations in Ca2þ homeostasis must lie upstream in the cascade of injury caused by Ab and that the oxidative injury might itself be Ca2þ dependent. Cells were protected by provision of GSH precursors (Abramov et al., 2004) or by upregulation of GSH synthesis (Barkats et al., 2000), strongly suggesting that GSH depletion plays a major role in the progression towards cell death. All these data point to a major role of oxidative stress in response to Ab and indicate multiple potential targets of oxidative damage. Several mechanisms have been proposed whereby Ab may increase ROS generation. Ab may even generate oxygen radicals directly in solution, but may also interact with a number of biological systems to increase the rate of radical production through modification or stimulation of intrinsic pathways. The most obvious of these are probably the activation of endogenous radical generating systems in microglia and possibly other cell types in the CNS by activation of the flavoprotein linked enzyme system NADPH oxidase. As indicated above, it has also been suggested that Ab may increase the production of ROS from mitochondria by causing damage to the mitochondrial respiratory chain. We have recenty found evidence for the expression of an NADPH oxidase in astrocytes, and have suggested that the activation of the astrocytic enzyme by Ab may play a central role in the pathophysiology of Ab toxicity (Abramov et al., 2004). We found using ROS sensitive dyes such as dihydroethidium or dicarboxyfluorescein, that Ab increases the rate of ROS generation in astrocytes (not in neurons) a response that is inhibited by diphenylene iodonium, DPI (Abramov et al., 2004). The rate of ROS increase was associated with GSH depletion, and both the increase in ROS and the loss of GSH were dependent on extracellular calcium, suggesting that the activation of the NADPH oxidase may be Ca2þ dependent and part of the [Ca2þ ]c response of astrocytes to the peptide (see above). These observations have raised a series of fascinating issues about the physiological role of the enzyme in astrocytes, its normal modes of activation and

430

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

regulation, its expression, and its contribution to pathological states, but all of these issues remain largely untested. Ab could also increase ROS generation through an action on mitochondria. I have discussed above evidence for oxidative damage to mitochondria in AD, including the failure of substrate supply through effects on a-KGDH and other enzymes of intermediary metabolism, but may also cause impaired activity of COX. The former will simply cause a loss of mitochondrial potential but will not increase mitochondrial ROS generation. However damage to COX may increase mitochondrial ROS generation as the intermediary complexes become fully reduced and transfer electrons to available oxygen if downstream carriers cannot accept them. This seems unlikely to represent a primary mechanism of cell injury in Ab toxicity–– after all the mitochondrial damage must come first and is likely to be a reflection of oxidative stress generated through activation of another pathway––such as the NADPH oxidase––but it will provide an amplifying mechanism that may compound oxidative injury initiated elsewhere. We found that Ab caused the slow progressive loss of mitochondrial membrane potential in astrocytes (Fig. 11A and B) but not in neurons. Superimposed on the the slow depolarization were abrupt and very large transient depolarisations which seemd to be clearly associated with the calcium transients described above (Fig. 11B). The slow loss of potential (but not the large transients) could be reversed through the provision of mitochondrial substrates, such as pyruvate, glutamate, TMPD/ascorbate or by antioxidants (Fig. 11C; Abramov et al., 2004). The entire mitochondrial response was blocked by low concentration sof diphenylene iodonium and apocynin, inhibitors of the NADPH oxidase. These observations strongly suggest that the mitochondria depolarize as a consequence of oxidative damage which impairs substrate supply perhaps through impaired glucose transport (Parpura-Gill et al., 1997; Alvarez et al., 2003). There have been suggestions that mitochondrial dysfunction may be a primary disorder in AD patients. Remarkably, mitochondria isolated from the platelets of AD patients and introduced into a q0 neuronal cell line by cell fusion showed increased rates of ROS generation and disturbed calcium balance compared to control cybrids (Sheehan et al., 1997; Swerdlow et al., 2000). The question raised by such observations is whether the mitochondrial dysfunction is primary or secondary, although it is hard to understand why platelet mitochondria should be affected by pathological events taking place in the CNS. The suggestions have been that AD might in fact result from the accumulation of mitochondrial mutations affecting COX. This notion clearly is beyond the scope of considerations of Ab toxicity but has certainly prompted some interesting investigations. What has been so consistent and surprising in our own studies (Abramov et al., 2003, 2004) has been the observation that all pathophysiology––changes in [Ca2þ ]c and changes in ROS generation––seem to be confined to astrocytes, and yet the cells that died later were predominantly the neurons. The astrocyte [Ca2þ ]c signals were not influenced at all by antioxidants. GSH was however depleted in both astrocytes and neurons and the GSH depletion was calcium dependent, strongly suggesting a sequence of events whereby Ab forms a channel in the astrocyte membrane, promoting calcium influx into astrocytes as a primary phenomenon. This then generates

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

431

Fig. 11. Effects of Amyloid beta peptide on mitochondria in astrocytes. (A) Application of Ab to a mixed culture of astrocytes and neurons causes a slow loss of mitochondrial potential in astrocytes only on which are superimposed abrupt large and reversible depolarisations. The images have been extracted from a series of images collected over time from a culture of astrocytes loaded with dequenching concentration of rhodamine 123 (rh 123). Note the appearance of bright cells in some fields and the gradual increase in the signal throughout the culture. (B) Here, [Ca2þ ]c and mitochondrial potential were measured simultaneously from astrocytes dual loaded with rh 123 and fura-2. Note the large transient mitochondrial depolarisation that clearly accompanies the [Ca2þ ]c signal. In the presence of additional mitochondrial substrate, in this case methyl succinate 5 mM (C), the slow loss of mitochondrial potential was abolished leaving the large transient depolarisations unaffected. The latter were abolished in the absence of extracellular calcium. (D) Diphenylene iodonium (DPI, 0.5 mM) an inhibitor of the NADPH oxidase, completely blocked the mitochondrial responses without having any impact on the [Ca2þ ]c signal (cells dual loaded with fura-2 and rh 123). These data were adapted from Abramov et al. (2004).

432

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

oxidative stress within the astrocytes, possibly through the activation of the NADPH oxidase, causing GSH depletion. Astrocyte GSH export can no longer sustain the requirements of neurons for GSH precursors and so the neurons also become GSH depleted. It is the experience of many investigators that depletion of neuronal GSH may be sufficient to cause neuronal death, as these cells seem much more vulnerable to endogenous pro-oxidants than the astrocytes, which appear far more robust. What emerges from these studies is that the toxicity of Ab is a neural response reflecting a complex series of interactions between glial and neuronal signals, involving glial mitochondrial oxidative stress, glial Ca2þ signals, impaired glialneuronal signaling and ultimately neuronal attrition. It has to be said also that I have been careful to discuss these issues in terms only of understanding Ab toxicity. It may be a long step from here to understanding what happens in the brain of a subject with AD. Nevertheless, one of the more exciting findings in our recently published work was the observation that the drug, clioquinol, an antimalarial which has been shown to confer some clinical benefit in animal models and recently in patients also abolished the calcium signals and the oxidative stress in astrocytes 11.5. Huntington’s disease Huntington’s disease (HD) is caused by an expansion of exonic CAG triplet repeats in the gene encoding the protein known as Huntingtin (Htt). Panov et al., 2002) found that mitochondria isolated from lymphoblasts of patients with HD had a lower membrane potential than mitochondria from control subjects. They also found that upon calcium addition, the mitochondria depolarised in response to smaller calcium loads than mitochondria from controls. They found a similar response in mitochondria extracted from the brains of transgenic mice which express the full-length mutant protein, huntingtin. Interestingly, this defect preceded the onset of pathological or behavioral abnormalities by months. The mutant huntingtin could be seen on neuronal mitochondrial membranes using electron microscopy with imunolabelling. Further and remarkably, incubation of normal mitochondria with a fusion protein containing an abnormally long polyglutamine repeat reproduced the mitochondrial calcium defect seen in human patients and transgenic animals. The authors suggested that, mitochondrial calcium abnormalities occur early in HD pathogenesis and may be a direct effect of mutant huntingtin on the organelle. A more recent study by Brustovetsky et al. (2003) also found that mitochondria from striatal neurons showed a decreased threshold for mPTP opening upon addition of calcium compared to mitochondria extracted from other areas of the brain, and suggested that this might confer greater vulnerability on these neurons in Huntington’s disease if polyglutamine repeats are seen in all tissues. 11.6. Parkinson’s disease There is a considerable body of evidence that links abnormalities of mitochondrial Complex I with the selective degeneration of nigro-striatal neurons in Parkinson’s disease. The most impressive model of the disease, that of MPTP (1-methyl-phenyl-

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

433

4-phenyl-1,2,3,6,-tetrahydropyridine) toxicity experienced by drug users and later widely used as an experimental model, is associated with damage to complex I of the mitochondrial respiratory chain by the derivative MPPþ (Langston and Ballard, 1983). This model is open to criticism, in that the conversion of MPTP to MPP+ requires monoamine oxidase which is found in the dopaminergic cells of the striatum. MPP+ further is a substrate for the DA transporter and so is accumulated selectively within striatal neurons where it is toxic. Thus there is no real mystery that the drug causes selective loss of these neurons, nor does it really give any clue as to the ‘normal’ mechanism of selective loss of these neurons in the disease. However, Complex I deficiency has also been widely reported in the substantia nigra of patients with sporadic PD (Parker et al., 1989; Schapira et al., 1989; Mizuno et al., 1989; Bindoff et al., 1989). It has even been suggested on epidemiological grounds that at least in some patients, the disease my reflect environmental exposure to the insecticide rotenone, the classical inhibitor of complex I. What does seem extraordinary is that the systemic administration of low doses of rotenone to rats seems to cause a PD like syndrome which includes the selective degeneration nigrostriatal neurons and the appearance of synuclein rich inclusion bodies which are a feature of the disease (Betarbet et al., 2000; Sherer et al., 2002). The specific site of action of rotenone seems unambiguously defined as mitochondrial complex I––the drug is relatively selective, and the same group remarkably also managed to transfect into cells a yeast subunit of complex I which lacks the rotenone binding site, and cells became resistant to rotenone (Sherer et al., 2003)––although of course this tells us little about the action of the drug in the whole animal. It seems that cell injury is not ‘simply’ a function of metabolic insufficiency, as similar degrees of ATP depletion induced by other poisons (inhibition of glycolysis with 2-deoxyglucose) failed to cause the same injury. Rather, cells could be protected with antioxidants, suggesting that the primary mechanism of injury is through oxidative stress. Complex I is particularly vulnerable to modification by oxidative stress, and in turn is a potential generator of ROS. What seems so extraordinary to me is that given systemic administration of a poison that targets complex I in all tissues, why should the striatal neurons be so selectively vulnerable? What happens in the heart? In other parts of the brain? The greatest puzzle in almost all of the neurodegenerative diseases is the way in which a widespread pathology can cause selective degeneration in a subset of neurons––the mSOD defect that only affects motoneurons, rotenone toxicity that only affects striatal neurons, the polyglutamine repeats that cause degeneration of basal ganglionic neurons and so on. In all of these cases, we seem to see some interplay between oxidative stress and mitochondrial dysfunction, but in no case do we have a really satisfactory answer to explain the selectivity of the disease phenotype.

12. In the heart: mitochondrial calcium overload and reperfusion injury Perhaps it seems self-evident that, if cells are deprived of glucose and oxygen, they will eventually die. In stroke and in a cardiac infarct, this clearly defines cell death in the short term. The factors that determine the rate of ATP depletion following

434

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

complete ischaemia are not well defined and seem important, as these will define the time available before irreversible damage will ensue. What are the major ATP consumers under these conditions? Studies in isolated cardiomyocytes suggest that the mitochondrial ATPase itself may be one of the major ATP consumers during ischaemia. Thus, in the absence of oxygen, mitochondrial respiration will stop. Mitochondrial potential will inevitably begin to dissipate, at a rate determined presumably by proton leak. In response, the equilibrium of the ATP synthase will push the enzyme to run in ‘reverse’ as a proton pumping ATPase, consuming ATP, translocating protons out of the mitochondria and acting as a brake on the rate of mitochondrial depolarisation. This can be seen directly in intact cells, in which the mitochondrial membrane potential is maintained well until intracellular ATP is depleted (e.g. see Duchen et al., 1993a,b) and the cell goes into a state of rigor contraction. Inhibition of the ATPase allows rapid dissipation of mitochondrial potential but slows the rate of ATP depletion during experimental ischaemia (see Leyssens et al., 1996). Interestingly, an endogenous protein, named IF-1, binds to the ATPase and inhibits ATP consumption in the hearts of larger mammals (Rouslin, 1991). The crystal structure and the mechanism of inhibition of the ATPase are well characterised (see Stock et al., 1999), but much less is known of the physiological properties of the protein in intact cells. Even less is known of IF1 function in tissues other than the heart. While cardiomyocytes are ischaemic, ATP falls progressively. Eventually, cells will die. However, there is another strange phenomenon, in which cells that have been able to withstand a period of ischaemia paradoxically die when they are reperfused. This has been called reperfusion injury or the oxygen paradox. Why should cells die when oxygen is reintroduced? There have been various suggestions. The simplest is the possibility that the mechanical changes as cells shorten to a hypercontracture when ATP is returned and cause membrane damage. However, consider the following scheme: as mitochondrial respiration is inhibited and ATP gradually falls, anaerobic glycolysis generates an intracellular acidosis. The acidification drives Naþ /Hþ exchange, raising intracellular Naþ , which in turn drives the plasmalemmal Naþ /Ca2þ exchanger to bring Ca2þ into the cell. Probably, as long as mitochondrial potential remains polarised, mitochondria will also become Ca2þ loaded, although this has not been demonstrated very convincingly. Indeed, it has also been suggested that mitochondria may become calcium loaded by reversal of the mitochondrial Naþ /Ca2þ exchanger (Griffiths, 1999), although thermodynamically this would presumably require a very large increase in intramitochondrial Naþ which has not yet been demonstrated. Nevertheless, mitochondrial potential will also gradually dissipate, itself serving to limit mitochondrial calcium accumulation. However, at reperfusion, we have a scenario when mitochondrial potential will abruptly recover as mitochondria begin to respire again, and the repolarisation occurs at a time in which cytosolic [Ca2þ ] is high. This will cause an immediate loading of mitochondria with calcium. It has also been suggested that the return to mitochondrial respiration after a period of anoxia may be accompanied by a burst of mitochondrial free radical generation. The combination of a high mitochondrial calcium concentration and oxidative stress, coupled with ATP depletion and a high Pi provides exactly the condition that

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

435

favours opening of the mitochondrial permeability transition pore (see above: ATP inhibits mPTP opening). This will collapse potential completely, promote rapid ATP loss and seems inevitably to be a point of no return. Studies of reperfusion injury probably provide the most convincing case of for a role of the mPTP in a defined pathological context. Thus, CsA is cardioprotective if introduced at the time of reperfusion but not before or afterwards, reperfusion is associated with the increased accumulation of intramitochondrial labelled 2-deoxyglucose (‘Hot DOG’; see Griffiths and Halestrap, 1995; for review see Halestrap, 1999). We have also recently shown that a newer inhibitor of mPTP opening, sanglifehrin, is protective at reperfusion but not before or after, and have shown that it inhibits mitochondrial depolarisation in cardiomyocytes induced by illumination of mitochondrially localised fluorescent dyes, confirming this phototoxicity as a model of mPTP opening (Hausenloy et al., 2003). 12.1. The ‘mitochondrial KATP channel’: a target for cytoprotection This is a complex story but one that has puzzled, confused and intrigued over recent years. Essentially, as I understand it, it was thought that opening ATP dependent Kþ channels in the heart might prove cardioprotective on the grounds that these channels, normally opened as [ATP] falls, serve to shorten the cardiac action potential and so would be expected to limit calcium influx. A group of drugs which increase the open probability of these channels, the Kþ channel openers or KCOs did indeed prove to be cardioprotective. However, it turned out that the pharmacological profile of these drugs did not match the efficacy at opening plasmalemmal K channels. Indeed, the most effective cardioprotective agents, BMS180448 and diazoxide, seemed to have almost no effect on plasmalemmal K channels (Grover et al., 1995a,b,c). Around the same time, several reports had appeared describing a mitochondrially expressed ATP dependent K channel (Inoue et al., 1991; Garlid et al., 1996). It turned out that the pharmacological profile of KCO efficacy matched that of the mitochondrial ATP dependent K channel (mKATP ), leading, not unreasonably to the suggestion that the cardioprotective action of these drugs must somehow involve an action at the mitochondrial channel. At present, as far as I am aware, no-one really understands how the mKATP channel really operates physiologically, nor what conditions will normally cause it to open. What is also most intriguing are the parallels between protection by KCOs and by ischaemic preconditioning (for a recent review, see Yellon and Downey, 2003). This describes a phenomenon whereby a relatively brief, sublethal episode of ischaemia initiates a series of signalling pathways that confers protection against injury by subsequent ischaemia. What seems so striking is that cardiac preconditioning is blocked by KATP channel blockers such as glibenclamide. The pharmacological profile of the channel blockers that prevent preconditioning and the channel openers that mimic it is consistent with a role of the mKATP channel rather than the plasmalemmal channel in this process. The role of PKC is also well established, but the relationship between these is not at all clear––does PKC perhaps phosphorylate a mitochondrial target to confer protection? Or does a change in mitochondrial

436

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

function initiate signaling mechanisms that activate PKC that then phosphorylates another target? The problems with a model that assumes a role for a mitochondrial Kþ channel are many, however. Opening a Kþ channel in the mitochondrial inner membrane will be expected to depolarise the mitochondrial potential, as Kþ is thought to be equally distributed across the membrane. It is not clear how mitochondrial depolarisation would be cardioprotective. One possibility is that a modest depolarisation might serve to limit mitochondrial calcium accumulation and therefore retard mitochondrial calcium overload and resultant damage (see above; Murata et al., 2001). This fits with some of our own observations in another model of cardioprotection, using NO. We found that exposure of cardiomyocytes to NO donors for a period of a few hours followed by washout conferred a significant protection to subsequent exposure to ischemic conditions (Rakhit et al., 2001). This was clearly associated with a modest mitochondrial depolarisation and a significant reduction in the capacity of mitochondria to accumulate calcium from a given local calcium load. The first evidence that diazoxide application actually directly affects mitochondrial function rather than acting elsewhere in the cell (i.e. at the plasma membrane), came from measurements of flavoprotein autofluorescence (Liu et al., 1998). O’Rourke’s group showed that diazoxide causes massive increases in the flavoprotein signal, indicating flavin oxidation, of a magnitude equivalent to that seen with a mitochondrial uncoupler. The basis for this response, adopted by the authors as a signature with which to ‘assay the activation of mKatp channels’ has never been very clear. One might expect an oxidation of intermediates if the drug were causing a mitochondrial depolarisation, but no depolarisation has been measured except at concentrations of diazoxide much higher than those required to cause protection. We have also tried to replicate these responses and even in islolated cardiomyocytes in which diazoxide is clearly protective, we see no significant changes in flavoprotein or NADH autofluorescence, suggesting that the mechanism of protection may be dissociated from whatever mechanism causes the autofluorescence changes (unpublished observations, but see also Lawrence et al. (2001)––who actually published the finding!). The picture is complicated even further as diazoxide is able to have effects on mitochondria quite independently of any action on mKATP channels (Grimmsmann and Rustenbeck, 1998; Ovide-Bordeaux et al., 2000), apparently acting as an inhibitor of mitochondrial Complex II, succinate dehydrogenase. Indeed, the every existence of mKATP channels has been challenged (Das et al., 2003). Further, other drugs, (such as 3-nitroproprionic acid) which acts as an inhibitor of Complex II, may also cause some kind of preconditioning in some models. Does this mean that all actions of diazoxide are due to actions on Complex II and that the mKATP channel is irrelevant? It seems to me that this remains an open question. The pharmacological tools available to investigate the channel remain inadequate––it is blocked by 5-hydroxydecanoate, (5-HD) but this also has curious metabolic effects and is not a good pharmacological tool. The bulk of this discussion dealt with the heart, perhaps because cardiac reperfusion injury is such a well established model, perhaps because the initial impetus for these studies came from the expression in the heart of a plasmalemmal KATP channel.

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

437

There are data suggesting that these same drugs may be protective in other tissues. Thus, the KATP channel opener cromakalim is apparently neuroprotective in glutamate neuroxtoxity (Lauritzen et al., 1997; Reshef et al., 1998), although the mechanism for this protection was not clear and a mitochondrial target was not suggested. The latter study suggested a very similar process for ischaemic preconditioning in neurons as that described in the heart involving upregulation of PKC and a process that is enhanced by K channel openers and blocked by their antagonists. The most recent literature suggests that perhaps the mKATP channel openers operate by modulating the open probability of the mPTP (Hausenloy et al., 2004). It seems clear that in cardiac cells treated with diazoxide the mPTP is less likely to open given a similar oxidative stress. The mechanism for this is not clear––is it that the mitochondria become less calcium loaded because perhaps they are somewhat depolarised? Does this really have anything at all to do with a mitochondrial Kþ channel? If it does not, what is the functional role of the mitochondrial Kþ channel? These questions remain wide open at the time of writing, but it is an active field which will hopefully develop rapidly as it clearly provides an exciting potential therapeutic approach to provide profound protection to compromised heart and possibly other tissues. What emerges from these very different studies––of mechanisms of neurodegeneration and of cardiac reperfusion injury is a common theme which seems to be reprised in many different forms of pathological process, in which changes in calcium signals and oxidative stress – however generated––converge on the mitochondrion. There, mitochondrial damage may result, with consequences either causing cellular dysfunction, which may prove to be reversible, or triggering changes that result in sufficient injury to cause cell death, either through ATP depletion and necrosis or by initiating the mitochondrial pathways to apoptotic cell death. It seems clear enough that if we can understand these mechanisms then it should become possible to develop rational therapeutic targets for a wide variety of different pathological states.

13. Postscript What I have attempted to do here is to discuss what I can of the current status of mitochondrial biology, attempting to highlight those aspects of modern mitochondrial studies that seem to me to be of particular potential interest in relation to disease. I firmly believe that it is necessary to understand the basic biological principles first before we have any chance of understanding how abnormalities might have an impact of cell or tissue function to cause the strange and often unpredictable consequences that we see in disease. It also seems to me self-evident that these structures play such a fundamental and complex role in cell biology that it is inevitable that disorders of almost any aspect of mitochondrial function––biogenesis, replication, fusion, fission, proton pumping, ROS generation, calcium handling, and, of course, ATP generation––will inevitably cause cellular dysfunction manifest as disease. The mystery remains how tissue or cell type specificity occurs, how a systemic disorder of one mitochondrial complex can cause a selective disease phenotype

438

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

such as optic nerve atrophy in Leber’s hereditary optic neuropathy while leaving other tissues intact, why only motoneurons die in ALS, or basal ganglia neurons in Parkinson’s disease. These puzzles await solutions and drive us on to devise new approaches to try to understand the secret world of the mitochondrion.

Acknowledgements Work in my laboratory has been funded by the Wellcome trust, the Medical Research Council and the Royal Society, to whom I am immensely grateful. I also thank my colleagues in the lab over recent years, whose industry has contributed to the work that I have described here––Ann Leyssens, Jake Jacobson, Julie Keelan, Olga Vergun, Andrey Abramov, Matt McKenzie, Olga Beskina. In particular I thank Dr. Laura Canevari, without whose support and encouragement none of this would have been written. The review has been embellished with Figures contributed generously by friends and colleagues who have been without fail most gracious in their willingness to contribute. For this I thank Drs. Guy Perkins (Department of Neurosciences and National Center for Microscopy and Imaging Research, University of California San Diego, La Jolla, USA), Manuel Rojo (INSERM U582, Paris France), Gyuri Szabadkai and Mariusz Wieckowski (Department of General Pathology, University of Ferrara, Ferrara Italy), Elizabeth Jonas (Department of Pharmacology, Yale University School of Medicine, New Haven, USA), and Hans Spelbrink (Institute of Medical Technology and Tampere University Hospital, University of Tampere, Tampere, Finland). I have placed some additional material on my web site at http://www.physiol.ucl.ac.uk/research/duchen_m/ which includes some animated sequences kindly provided by Dr. Mariusz Wieckowski.

References Aarts, M., Liu, Y., Liu, L., Besshoh, S., Arundine, M., Gurd, J.W., Wang, Y.T., Salter, M.W., Tymianski, M., 2002. Treatment of ischemic brain damage by perturbing NMDA receptor-PSD-95 protein interactions. Science 298 (5594), 846–850. Abramov, A.Y., Canevari, L., Duchen, M.R., 2003. Changes in intracellular calcium and glutathione in astrocytes as the primary mechanism of amyloid neurotoxicity. J. Neurosci. 23 (12), 5088–5095. Abramov, A.Y., Canevari, L., Duchen, M.R., 2004. Beta-amyloid peptides induce mitochondrial dysfunction and oxidative stress in astrocytes and death of neurons through activation of NADPH oxidase. J. Neurosci. 24 (2), 565–575. Ainscow, E.K., Rutter, G.A., 2001. Mitochondrial priming modifies Ca2þ oscillations and insulin secretion in pancreatic islets. Biochem. J. 353 (Pt 2), 175–180. Alvarez, G., Ramos, M., Ruiz, F., Satrustegui, J., Bogonez, E., 2003. Pyruvate protection against betaamyloid-induced neuronal death: role of mitochondrial redox state. J. Neurosci. Res. 73 (2), 260–269. Anandatheerthavarada, H.K., Biswas, G., Robin, M.A., Avadhani, N.G., 2003. Mitochondrial targeting and a novel transmembrane arrest of Alzheimer’s amyloid precursor protein impairs mitochondrial function in neuronal cells. J. Cell. Biol. 161 (1), 41–54. Anthony, G., Stroh, A., Lottspeich, F., Kadenbach, B., 1990. Different isozymes of cytochrome c oxidase are expressed in bovine smooth muscle and skeletal or heart muscle. Febs Lett. 277, 97–100.

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

439

Archer, S., Michelakis, E., 2002. The mechanism(s) of hypoxic pulmonary vasoconstriction: potassium channels, redox O(2) sensors, and controversies. News Physiol. Sci. 17, 131–137. Arispe, N., Doh, M., 2002. Plasma membrane cholesterol controls the cytotoxicity of Alzheimer’s disease Ab P (1-40) and (1-42) peptides. FASEB J. 16, 1526–1536. Arispe, N., Pollard, H.B., Rojas, E., 1993. Giant multilevel cation channels formed by Alzheimer disease amyloid b-protein [AbP-(1-40)] in bilayer membranes. Proc. Natl. Acad. Sci. USA 90, 10573–10577. Arispe, N., Pollard, H.B., Rojas, E., 1996. Zn2þ interaction with Alzheimer amyloid b-protein calcium channels. Proc. Natl. Acad. Sci. USA 93, 1710–1715. Arnold, S., Kadenbach, B., 1997. Cell respiration is controlled by ATP, an allosteric inhibitor of cytochrome-c oxidase. Eur. J. Biochem. 249, 350–354. Babcock, D.F., Herrington, J., Goodwin, P.C., Park, Y.B., Hille, B., 1997. Mitochondrial participation in the intracellular Ca2þ network. J. Cell Biol. 136 (4), 833–844. Barkats, M., Millecamps, S., Abrioux, P., Geoffroy, M.C., Mallet, J., 2000. Overexpression of glutathione peroxidase increases the resistance of neuronal cells to Ab-mediated neurotoxicity. J. Neurochem. 75, 1438–1446. Beal, M.F., 1998. Mitochondrial dysfunction in neurodegenerative diseases. Biochim. Biophys. Acta. 1366 (1–2), 211–223. Beltran, B., Mathur, A., Duchen, M.R., Erusalimsky, J.D., Moncada, S., 2000. The effect of nitric oxide on cell respiration: A key to understanding its role in cell survival or death. PNAS (USA) 97, 14602–14607. Bernardi, P., Azzone, G.F., 1981. Cytochrome c as an electron shuttle between the outer and inner mitochondrial membranes. J. Biol. Chem. 256 (14), 7187–7192. Bernardi, P., Azzone, G.F., 1982. A membrane potential-modulated pathway for Ca2þ efflux in rat liver mitochondria. FEBS Lett. 139, 13–16. Betarbet, R., Sherer, T.B., MacKenzie, G., Garcia-Osuna, M., Panov, A.V., Greenamyre, J.T., 2000. Chronic systemic pesticide exposure reproduces features of Parkinson’s disease. Nat. Neurosci. 3 (12), 1301–1306. Beutner, G., Sharma, V.K., Giovannucci, D.R., Yule, D.I., Sheu, S.S., 2001. Identification of a ryanodine receptor in rat heart mitochondria. J. Biol. Chem. 276 (24), 21482–21488. Billups, B., Forsythe, I.D., 2002. Presynaptic mitochondrial calcium sequestration influences transmission at mammalian central synapses. J. Neurosci. 22 (14), 5840–5847. Bindoff, L.A., Birch-Machin, M., Cartlidge, N.E., Parker Jr., W.D., Turnbull, D.M., 1989. Mitochondrial function in Parkinson’s disease. Lancet 2 (8653), 49. Boitier, E., Rea, R., Duchen, M.R., 1999. Mitochondria exert a negative feedback on the propagation of intracellular Ca2þ waves in rat cortical astrocytes. J. Cell Biol. 145 (4), 795–808. Borutaite, V., Matthias, A., Harris, H., Moncada, S., Brown, G.C., 2001. Reversible inhibition of cellular respiration by nitric oxide in vascular inflammation. Am. J. Physiol. Heart Circ. Physiol. 281 (6), H2256–H2260. Bossy-Wetzel, E., Barsoum, M.J., Godzik, A., Schwarzenbacher, R., Lipton, S.A., 2003. Mitochondrial fission in apoptosis, neurodegeneration and aging. Curr. Opin. Cell Biol. 15 (6), 706–716. Bossy-Wetzel, E., Newmeyer, D.D., Green, D.R., 1998. Mitochondrial cytochrome c release in apoptosis occurs upstream of DEVD-specific caspase activation and independently of mitochondrial transmembrane depolarization. EMBO J. 17 (1), 37–49. Brand, M.D., 1985. The stoichiometry of the exchange catalysed by the mitochondrial calcium/sodium antiporter. Biochem. J. 229, 161–166. Brealey, D., Singer, M., 2003. Mitochondrial dysfunction in sepsis. Curr. Infect. Dis. Rep. 5 (5), 365– 371. Brini, M., Pinton, P., King, M.P., Davidson, M., Schon, E.A., Rizzuto, R.A., 1999. Calcium signaling defect in the pathogenesis of a mitochondrial DNA inherited oxidative phosphorylation deficiency. Nat. Med. 5 (8), 951–954. Brustovetsky, N., Brustovetsky, T., Purl, K.J., Capano, M., Crompton, M., Dubinsky, J.M., 2003. Increased susceptibility of striatal mitochondria to calcium-induced permeability transition. J. Neurosci. 23 (12), 4858–4867. Brustovetsky, N., Dubinsky, J.M., 2000. Dual responses of CNS mitochondria to elevated calcium. J. Neurosci. 20 (1), 103–113.

440

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

Buckler, K.J., Vaughan_Jones, R.D., 1998. Effects of mitochondrial uncouplers on intracellular calcium, pH and membrane potential in rat carotid body type I cells. J. Physiol. 513 (Pt 3), 819–833. Buckman, J.F., Reynolds, I.J., 2001. Spontaneous changes in mitochondrial membrane potential in cultured neurons. J. Neurosci. 21 (14), 5054–5065. Buntinas, L., Gunter, K.K., Sparagna, G.C., Gunter, T.E., 2001. The rapid mode of calcium uptake into heart mitochondria (RaM): comparison to RaM in liver mitochondria. Biochim. Biophys. Acta 1504, 248–261. Bunting, J.R., 1992. A test of the singlet oxygen mechanism of cationic dye photosensitization of mitochondrial damage. Photochem. Photobiol. 55, 81–87. Bush, A.I., 2003. The metallobiology of Alzheimer’s disease. Trends Neurosci. 26, 207–214. Cabezon, E., Montgomery, M.G., Leslie, A.G., Walker, J.E., 2003. The structure of bovine F1-ATPase in complex with its regulatory protein IF1. Nat. Struct. Biol. 10 (9), 744–750. Canevari, L., Clark, J.B., Bates, T.E., 1999. b-amyloid fragment 25–35 selectively decreases complex IV activity in isolated mitochondria. FEBS Lett. 457, 131–134. Carriedo, S.G., Sensi, S.L., Yin, H.Z., Weiss, J.H., 2000. AMPA exposures induce mitochondrial Ca(2+) overload and ROS generation in spinal motor neurons in vitro. J. Neurosci. 20 (1), 240–250. Carriedo, S.G., Yin, H.Z., Weiss, J.H., 1996. Motor neurons are selectively vulnerable to AMPA/kainate receptor-mediated injury in vitro. J. Neurosci. 16 (13), 4069–4079. Casley, C.S., Canevari, L., Land, J.M., Clark, J.B., Sharpe, M.A., 2002. b-amyloid inhibits integrated mitochondrial respiration and key enzyme activities. J. Neurochem. 80, 91–100. Castilho, R.F., Ward, M.W., Nicholls, D.G., 1999. Oxidative stress, mitochondrial function, and acute glutamate excitotoxicity in cultured cerebellar granule cells. J. Neurochem. 72 (4), 1394–1401. Cecchi, C., Latorraca, S., Sorbi, S., Iantomasi, T., Favilli, F., Vincenzini, M.T., Liguri, G., 1999. Gluthatione level is altered in lymphoblasts from patients with familial Alzheimer’s disease. Neurosci. Lett. 275 (2), 152–154. Chance, B., Baltscheffsky, H., 1958. Respiratory enzymes in oxidative phosphorylation. VII. Binding of intramitochondrial reduced pyridine nucleotide. J. Biol. Chem. 233 (3), 736–739. Chance, B., Schoener, B., Oshino, R., Itshak, F., Nakase, Y., 1979. Oxidation–reduction ratio studies of mitochondria in freeze-trapped samples. NADH and flavoprotein fluorescence signals. J. Biol. Chem. 254, 4764–4771. Chapman, P.F., White, G.L., Jones, M.W., Cooper-Blacketer, D., Marshall, V.J., Irizarry, M., Younkin, L., Good, M.A., Bliss, T.V.P., Hyman, B.T., Younkin, S.G., Hsiao, K.K., 1999. Impaired synaptic plasticity and learning in aged amyloid precursor protein transgenic mice. Nat. Neurosci. 2, 271– 276. Chen, Q., Vazquez, E.J., Moghaddas, S., Hoppel, C.L., Lesnefsky, E.J., 2003. Production of reactive oxygen species by mitochondria: central role of complex III. J. Biol. Chem. 278 (38), 36027–36031. Chernyak, B.V., 1997. Cyclosporin A-sensitive release of Ca2þ from mitochondria in intact thymocytes. FEBS Lett. 418 (1–2), 131–134. Choi, D.W., 1992. Excitotoxic cell death. J. Neurobiol. 23 (9), 1261–1276. Clapham, J.C., Arch, J.R., Chapman, H., Haynes, A., Lister, C., Moore, G.B., Piercy, V., Carter, S.A., Lehner, I., Smith, S.A., Beeley, L.J., Godden, R.J., Herrity, N., Skehel, M., Changani, K.K., Hockings, P.D., Reid, D.G., Squires, S.M., Hatcher, J., Trail, B., Latcham, J., Rastan, S., Harper, A.J., Cadenas, S., Buckingham, J.A., Brand, M.D., Abuin, A., 2000. Mice overexpressing human uncoupling protein-3in skeletal muscle are hyperphagic and lean. Nature 406 (6794), 415–418. Clarke, S.J., McStay, G.P., Halestrap, A.P., 2002. Sanglifehrin A acts as a potent inhibitor of the mitochondrial permeability transition and reperfusion injury of the heart by binding to cyclophilin-D at a different site from cyclosporin A. J. Biol. Chem. 277 (38), 34793–34799. Clement, A.M., Nguyen, M.D., Roberts, E.A., Garcia, M.L., Boillee, S., Rule, M., McMahon, A.P., Doucette, W., Siwek, D., Ferrante, R.J., Brown Jr., R.H., Julien, J.P., Goldstein, L.S., Cleveland, D.W., 2003. Wild-type nonneuronal cells extend survival of SOD1 mutant motor neurons in ALS mice. Science 302 (5642), 113–117. Collins, T.J., Berridge, M.J., Lipp, P., Bootman, M.D., 2002. Mitochondria are morphologically and functionally heterogeneous within cells. Embo J. 21, 1616–1627.

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

441

Collins, T.J., Bootman, M.D., 2003. Mitochondria are morphologically heterogeneous within cells. J. Exp. Biol. 206, 1993–2000. Cormack, B.P., Valdivia, R.H., Falkow, S., 1996. FACS-optimized mutants of the green fluorescent protein (GFP). Gene 173, 33–38. Crompton, M., 1999. The mitochondrial permeability transition pore and its role in cell death. Biochem. J. 341 (Pt 2), 233–249. Crompton, M., 2000. Mitochondrial intermembrane junctional complexes and their role in cell death. J. Physiol. 529 (Pt 1), 11–21. Crompton, M., Andreeva, L., 1994. On the interactions of Ca2þ and cyclosporin A with a mitochondrial inner membrane pore: a study using cobaltammine complex inhibitors of the Ca2þ uniporter. Biochem. J. 302 (Pt 1), 181–185. Crompton, M., Virji, S., Doyle, V., Johnson, N., Ward, J.M., 1999. The mitochondrial permeability transition pore. Biochem. Soc. Symp. 66, 167–179. Csordas, G., Thomas, A.P., Hajnoczky, G., 1999. Quasi-synaptic calcium signal transmission between endoplasmic reticulum and mitochondria. EMBO J. 18 (1), 96–108. Cubitt, A.B., Heim, R., Adams, S.R., Boyd, A.E., Gross, L.A., Tsien, R.Y., 1995. Understanding, improving and using green fluorescent proteins. Trends Biochem. Sci. 20, 448–455. Das, M., Parker, J.E., Halestrap, A.P., 2003. Matrix volume measurements challenge the existence of diazoxide/glibencamide-sensitive KATP channels in rat mitochondria. J. Physiol. 547 (Pt 3), 893–902. Davis, A.F., Clayton, D.A., 1996. In situ localization of mitochondrial DNA replication in intact mammalian cells. J. Cell Biol. 135 (4), 883–893. David, G., Barrett, J.N., Barrett, E.F., 1998. Evidence that mitochondria buffer physiological Ca2þ loads in lizard motor nerve terminals. J. Physiol. 509, 59–65. De Giorgi, F., Lartigue, L., Bauer, M.K., Schubert, A., Grimm, S., Hanson, G.T., Remington, S.J., Youle, R.J., Ichas, F., 2002. The permeability transition pore signals apoptosis by directing Bax translocation and multimerization. FASEB J. 16 (6), 607–609. Dedkova, E.N., Ji, X., Lipsius, S.L., Blatter, L.A., 2004. Mitochondrial calcium uptake stimulates nitric oxide production in mitochondria of bovine vascular endothelial cells. Am. J. Physiol. Cell Physiol. 286 (2), C406–C415. Dekkers, J., Bayley, P., Dick, J.R., Schwaller, B., Berchtold, M.W., Greensmith, L., 2004. Over-expression of parvalbumin in transgenic mice rescues motoneurons from injury-induced cell death. Neuroscience 123 (2), 459–466. de la Monte, S.M., Luong, T., Neely, T.R., Robinson, D., Wands, J.R., 2000. Mitochondrial DNA damage as a mechanism of cell loss in Alzheimer’s disease. Lab. Invest. 80, 1323–1335. Deng, Y., Marko, M., Buttle, K.F., Leith, A., Mieczkowski, M., Mannella, C.A., 1999. Cubic membrane structure in amoeba (Chaos carolinensis) mitochondria determined by electron microscopictomography. J. Struct. Biol. 127, 231–239. Dringen, R., Hirrlinger, J., 2003. Glutathione pathways in the brain. Biol. Chem. 384, 505–516. Duchen, M.R., 1992. Ca2þ -dependent changes in the mitochondrial energetics in single dissociated mouse sensory neurons. Biochem. J. 283, 41–50. Duchen, M.R., 1999. Contributions of mitochondria to animal physiology: from homeostatic sensor to calcium signalling and cell death. J. Physiol. 516 (Pt 1), 1–17. Duchen, M.R., 2000. Mitochondria and Ca2þ in cell physiology and pathophysiology. Cell Calcium 28 (5–6), 339–348. Duchen, M.R., Biscoe, T.J., 1992a. Relative mitochondrial membrane potential and [Ca2þ ]i in type I cells isolated from the rabbit carotid body. J. Physiol. 450, 33–61. Duchen, M.R., Biscoe, T.J., 1992b. Mitochondrial function in type I cells isolated from rabbit arterial chemoreceptors. J. Physiol. 450, 13–31. Duchen, M.R., Leyssens, A., Crompton, M., 1998. Transient mitochondrial depolarizations reflect focal sarcoplasmic reticular calcium release in single rat cardiomyocytes. J. Cell Biol. 142 (4), 975– 988. Duchen, M.R., McGuinness, O., Brown, L.A., Crompton, M., 1993a. On the involvement of a cyclosporin A sensitive mitochondrial pore in myocardial reperfusion injury. Cardiovasc. Res. 27 (10), 1790–1794.

442

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

Duchen, M.R., Smith, P.A., Ashcroft, F.M., 1993b. Substrate-dependent changes in mitochondrial function, intracellular free calcium concentration and membrane channels in pancreatic beta-cells. Biochem. J. 294 (Pt 1), 35–42. Duchen, M.R., Surin, A., Jacobson, J., 2003. Imaging mitochondrial function in intact cells. Methods Enzymol. 361, 353–389. Dugan, L.L., Sensi, S.L., Canzoniero, L.M., Handran, S.D., Rothman, S.M., Lin, T.S., Goldberg, M.P., Choi, D.W., 1995. Mitochondrial production of reactive oxygen species in cortical neurons following exposure to N-methyl-D-aspartate. J. Neurosci. 15 (10), 6377–6388. Duguez, S., Feasson, L., Denis, C., Freyssenet, D., 2002. Mitochondrial biogenesis during skeletal muscle regeneration. Am. J. Physiol. Endocrinol. Metab. 282, E802–E809. Dykens, J.A., 1994. Isolated cerebral and cerebellar mitochondria produce free radicals when exposed to elevated Ca2þ and Naþ : implications for neurodegeneration. J. Neurochem. 63 (2), 584–591. Echtay, K.S., Roussel, D., St-Pierre, J., Jekabsons, M.B., Cadenas, S., Stuart, J.A., Harper, J.A., Roebuck, S.J., Morrison, A., Pickering, S., Clapham, J.C., Brand, M.D., 2002a. Superoxide activates mitochondrial uncoupling proteins. Nature 415 (6867), 96–99. Echtay, K.S., Murphy, M.P., Smith, R.A., Talbot, D.A., Brand, M.D., 2002b. Superoxide activates mitochondrial uncoupling protein 2 from the matrix side. Studies using targeted antioxidants. J. Biol. Chem. 277 (49), 47129–47135. Eng, J., Lynch, R.M., Balaban, R.S., 1989. Nicotinamide adenine dinucleotide fluorescence spectroscopy and imaging of isolated cardiac myocytes. Biophys. J. 55, 621–630. Fabrizio, P., Pozza, F., Pletcher, S.D., Gendron, C.M., Longo, V.D., 2001. Regulation of longevity and stress resistance by Sch9 in yeast. Science 292 (5515), 288–290. Ferguson, Nicholls, 2002. Bioenergetics 3. Academic Press. Filippin, L., Magalhaes, P.J., Di Benedetto, G., Colella, M., Pozzan, T., 2003. Stable interactions between mitochondria and endoplasmic reticulum allow rapid accumulation of calcium in a subpopulation of mitochondria. J. Biol. Chem. 278 (40), 39224–39234. Fink, C., Morgan, F., Loew, L.M., 1998. Intracellular fluorescent probe concentrations by confocal microscopy. Biophys. J. 75, 1648–1658. Follmann, K., Arnold, S., Ferguson_Miller, S., Kadenbach, B., 1998. Cytochrome c oxidase from eucaryotes but not from procaryotes is allosterically inhibited by ATP. Biochem. Mol. Biol. Int. 45, 1047–1055. Frank, J., Wagenknecht, T., McEwen, B.F., Marko, M., Hsieh, C.E., Mannella, C.A., 2002. Threedimensional imaging of biological complexity. J. Struct. Biol. 138, 85–91. Frank, S., Gaume, B., Bergmann-Leitner, E.S., Leitner, W.W., Robert, E.G., Catez, F., Smith, C.L., Youle, R.J., 2001. The role of dynamin-related protein 1, a mediator of mitochondrial fission, in apoptosis. Dev. Cell 1 (4), 515–525. Frazier, A.E., Chacinska, A., Truscott, K.N., Guiard, B., Pfanner, N., Rehling, P., 2003. Mitochondria use different mechanisms for transport of multispanning membrane proteins through the intermembrane space. Mol. Cell Biol. 23 (21), 7818–7828. Frey, T.G., Mannella, C.A., 2000. The internal structure of mitochondria. Trends Biochem. Sci. 25, 319– 324. Friberg, H., Wieloch, T., 2002. Mitochondrial permeability transition in acute neurodegeneration. Biochimie 84 (2–3), 241–250. Ganfornina, M.D., Lopez_Barneo, J., 1991. Single K+ channels in membrane patches of arterial chemoreceptor cells are modulated by O2 tension. Proc. Natl. Acad. Sci. USA 88, 2927–2930. Garlid, K.D., Paucek, P., Yarov_Yarovoy, V., Sun, X., Schindler, P.A., 1996. The mitochondrial KATP channel as a receptor for potassium channel openers. J. Biol. Chem. 271, 8796–8799. Ghafourifar, P., Richter, C., 1997. Nitric oxide synthase activity in mitochondria. FEBS Lett. 418 (3), 291–296. Gibson, G.E., Park, L.C., Sheu, K.F., Blass, J.P., Calingasan, N.Y., 2000. The a-ketoglutarate dehydrogenase complex in neurodegeneration. Neurochem. Int. 36 (2), 97–112. Gibson, G.E., Zhang, H., Sheu, K.F., Bogdanovich, N., Lindsay, J.G., Lannfelt, L., Vestling, M., Cowburn, R.F., 1998. a-ketoglutarate dehydrogenase in Alzheimer brains bearing the APP670/671 mutation. Ann. Neurol. 44, 676–681.

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

443

Gilabert, J.A., Bakowski, D., Parekh, A.B., 2001. Energized mitochondria increase the dynamic range over which inositol 1,4,5-trisphosphate activates store-operated calcium influx. EMBO J. 20, 2672– 2679. Gincel, D., Zaid, H., Shoshan-Barmatz, V., 2001. Calcium binding and translocation by the voltagedependent anion channel: a possible regulatory mechanism in mitochondrial function. Biochem J. 358, 147–155. Giulivi, C., 2003. Characterization and function of mitochondrial nitric-oxide synthase. Free Radic. Biol. Med. 34 (4), 397–408. Green, K.N., Peers, C., 2001. Amyloid b peptides mediate hypoxic augmentation of Ca2þ channels. J. Neurochem. 77, 953–956. Griffiths, E.J., 1999. Reversal of mitochondrial Na/Ca exchange during metabolic inhibition in rat cardiomyocytes. FEBS Lett. 453 (3), 400–404. Griffiths, E.J., Halestrap, A.P., 1995. Mitochondrial non-specific pores remain closed during cardiac ischaemia, but open upon reperfusion. Biochem. J. 307 (Pt 1), 93–98. Grimmsmann, T., Rustenbeck, I., 1998. Direct effects of diazoxide on mitochondria in pancreatic B-cells and on isolated liver mitochondria. Br. J. Pharmacol. 123 (5), 781–788. Griparic, L., van_der_Bliek, A.M., 2001. The many shapes of mitochondrial membranes. Traffic 2, 235– 244. Grover, G.J., D_Alonzo, A.J., Hess, T., Sleph, P.G., Darbenzio, R.B., 1995a. Glyburide-reversible cardioprotective effect of BMS-180448 is independent of action potential shortening. Cardiovasc. Res. 30, 731–738. Grover, G.J., D_Alonzo, A.J., Parham, C.S., Darbenzio, R.B., 1995b. Cardioprotection with the KATP opener cromakalim is not correlated with ischemic myocardial action potential duration. J. Cardiovasc. Pharmacol. 26, 145–152. Grover, G.J., McCullough, J.R., D_Alonzo, A.J., Sargent, C.A., Atwal, K.S., 1995c. Cardioprotective profile of the cardiac-selective ATP-sensitive potassium channel opener BMS-180448. J. Cardiovasc. Pharmacol. 25, 40–50. Gunter, T.E., Buntinas, L., Sparagna, G., Eliseev, R., Gunter, K., 2000. Mitochondrial calcium transport: mechanisms and functions. Cell Calcium 28, 285–296. Hajnoczky, G., Hager, R., Thomas, A.P., 1999. Mitochondria suppress local feedback activation of inositol 1,4,5-trisphosphate receptors by Ca2þ . J. Biol. Chem. 274 (20), 14157–14162. Hajnoczky, G., Robb-Gaspers, L.D., Seitz, M.B., Thomas, A.P., 1995. Decoding of cytosolic calcium oscillations in the mitochondria. Cell 82, 415–424. Halestrap, A.P., 1999. The mitochondrial permeability transition: its molecular mechanism and role in reperfusion injury. Biochem. Soc. Symp. 66, 181–203. Harris, M.E., Wang, Y., Pedigo Jr., N.W., Hensley, K., Butterfield, D.A., Carney, J.M., 1996. Amyloid beta peptide (25–35) inhibits Na+-dependent glutamate uptake in rat hippocampal astrocyte cultures. J. Neurochem. 67 (1), 277–286. Hausenloy, D.J., Duchen, M.R., Yellon, D.M., 2003. Inhibiting mitochondrial permeability transition pore opening at reperfusion protects against ischaemia-reperfusion injury. Cardiovasc. Res. 60 (3), 617–625. Hausenloy, D., Wynne, A., Duchen, M., Yellon, D., 2004. Transient mitochondrial permeability transition pore opening mediates preconditioning-induced protection. Circulation, in press. Hirai, K., Aliev, G., Nunomura, A., Fujioka, H., Russell, R.L., Atwood, C.S., Johnson, A.B., Kress, Y., Vinters, H.V., Tabaton, M., Shimohama, S., Cash, A.D., Siedlak, S.L., Harris, P.L., Jones, P.K., Petersen, R.B., Perry, G., Smith, M.A., 2001. Mitochondrial abnormalities in Alzheimer’s disease. J. Neurosci. 21, 3017–3023. Ho, R., Ortiz, D., Shea, T.B., 2001. Amyloid-b promotes calcium influx and neurodegeneration via stimulation of L voltage-sensitive calcium channels rather than NMDA channels in cultured neurons. J. Alzheimer’s Dis. 3, 479–483. Hollenbeck, P.J., 1996. The pattern and mechanism of mitochondrial transport in axons. Front Biosci. 1, d91–d102. Horikawa, Y., Goel, A., Somlyo, A.P., Somlyo, A.V., 1998. Mitochondrial calcium in relaxed and tetanized myocardium. Biophys. J. 74, 1579–1590.

444

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

Hoth, M., Fanger, C.M., Lewis, R.S., 1997. Mitochondrial regulation of store-operated calcium signaling in T lymphocytes. J. Cell. Biol. 137 (3), 633–648. Hsieh, C.E., Marko, M., Frank, J., Mannella, C.A., 2002. Electron tomographic analysis of frozenhydrated tissue sections. J. Struct. Biol. 138, 63–73. Huang, H.M., Ou, H.C., Hsieh, S.J., 2000. Antioxidants prevent amyloid peptide-induced apoptosis and alteration of calcium homeostasis in cultured cortical neurons. Life Sci. 66, 1879–1892. Huser, J., Blatter, L.A., 1999. Fluctuations in mitochondrial membrane potential caused by repetitive gating of the permeability transition pore. Biochem. J. 343 (Pt 2), 311–317. Huser, J., Rechenmacher, C.E., Blatter, L.A., 1998. Imaging the permeability pore transition in single mitochondria. Biophys. J. 74 (4), 2129–2137. Ibanez, V., Pietrini, P., Alexander, G.E., Furey, M.L., Teichberg, D., Rajapakse, J.C., Rapoport, S.I., Schapiro, M.B., Horwitz, B., 1998. Regional glucose metabolic abnormalities are not the result of atrophy in Alzheimer’s disease. Neurology 50, 1585–1593. Ichas, F., Mazat, J.P., 1998. From calcium signaling to cell death: two conformations for the mitochondrial permeability transition pore. Switching from low- to high-conductance state. Biochim. Biophys. Acta 1366 (1–2), 33–50. Igbavboa, U., Pfeiffer, D.R., 1988. EGTA inhibits reverse uniport-dependent Ca2þ release from uncoupled mitochondria. Possible regulation of the Ca2þ uniporter by a Ca2þ binding site on the cytoplasmic side of the inner membrane. J. Biol. Chem. 263, 1405–1412. Inoue, I., Nagase, H., Kishi, K., Higuti, T., 1991. ATP-sensitive K+ channel in the mitochondrial inner membrane. Nature 352, 244–247. Inoue, K., Nakada, K., Ogura, A., Isobe, K., Goto, Y., Nonaka, I., Hayashi, J.I., 2000. Generation of mice with mitochondrial dysfunction by introducing mouse mtDNA carrying a deletion into zygotes. Nat. Genet. 26 (2), 176–181. Jacobson, J., Duchen, M.R., 2002. Mitochondrial oxidative stress and cell death in astrocytes–requirement for stored Ca2þ and sustained opening of the permeability transition pore. J. Cell Sci. 115 (Pt 6), 1175– 1188. Jonas, E.A., Buchanan, J., Kaczmarek, L.K., 1999. Prolonged activation of mitochondrial conductances during synaptic transmission. Science 286 (5443), 1347–1350. Jouaville, L.S., Ichas, F., Holmuhamedov, E.L., Camacho, P., Lechleiter, J.D., 1995. Synchronization of calcium waves by mitochondrial substrates in Xenopus laevis oocytes. Nature 377 (6548), 438–441. Jouaville, L.S., Pinton, P., Bastianutto, C., Rutter, G.A., Rizzuto, R.., 1999. Regulation of mitochondrial ATP synthesis by calcium: evidence for a long-term metabolic priming. Proc. Natl. Acad. Sci. USA 96 (24), 13807–13812. Jung, D.W., Baysal, K., Brierley, G.P., 1995. The sodium–calcium antiport of heart mitochondria is not electroneutral. J. Biol. Chem. 270, 672–678. Kadenbach, B., Huttemann, M., Arnold, S., Lee, I., Bender, E., 2000. Mitochondrial energy metabolism is regulated via nuclear-coded subunits of cytochrome c oxidase. Free Rad. Biol. Med. 29, 211–221. Kaftan, E.J., Xu, T., Abercrombie, R.F., Hille, B., 2000. Mitochondria shape hormonally induced cytoplasmic calcium oscillations and modulate exocytosis. J. Biol. Chem. 275, 25465–25470. Kanaji, S., Iwahashi, J., Kida, Y., Sakaguchi, M., Mihara, K., 2000. Characterization of the signal that directs Tom20 to the mitochondrial outer membrane. J. Cell Biol. 151, 277–288. Kapus, A., Szaszi, K., Kaldi, K., Ligeti, E., Fonyo, A., 1991. Is the mitochondrial Ca2þ uniporter a voltage-modulated transport pathway? FEBS Lett. 282 (1), 61–64. Karbowski, M., Arnoult, D., Chen, H., Chan, D.C., Smith, C.L., Youle, R.J., 2004. Quantitation of mitochondrial dynamics by photolabeling of individual organelles shows that mitochondrial fusion is blocked during the Bax activation phase of apoptosis. J. Cell Biol. 164 (4), 493–499. Kasparova, J., Lisa, V., Tucek, S., Dolezal, V., 2001. Chronic exposure of NG108-15 cells to amyloid b peptide (Ab142 ) abolishes calcium influx via N-type calcium channels. Neurochem. Res. 26, 1079– 1084. Kawahara, M., Kuroda, Y., 2001. Intracellular calcium changes in neuronal cells induced by Alzheimer’s b-amyloid protein are blocked by estradiol and cholesterol. Cell. Mol. Neurobiol. 21, 1–13. Keelan, J., Allen, N.J., Antcliffe, D., Pal, S., Duchen, M.R., 2001. Quantitative imaging of glutathione in hippocampal neurons and glia in culture using monochlorobimane. J. Neurosci. Res. 66, 873–884.

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

445

Keelan, J., Vergun, O., Duchen, M.R., 1999. Excitotoxic mitochondrial depolarisation requires both calcium and nitric oxide in rat hippocampal neurons. J. Physiol. 520 (Pt 3), 797–813. Kinnally, K.W., Lohret, T.A., Campo, M.L., Mannella, C.A., 1996. Perspectives on the mitochondrial multiple conductance channel. J. Bioenerg. Biomembr. 28 (2), 115–123. Kirichok, Y., Krapivinsky, G., Clapham, D.E., 2004. The mitochondrial calcium uniporter is a highly selective ion channel. Nature 427 (6972), 360–364. Kish, S.J., 1997. Brain energy metabolizing enzymes in Alzheimer’s disease: a-ketoglutarate dehydrogenase complex and cytochrome oxidase. Ann. N.Y. Acad. Sci. 826, 218–228. Kish, S.J., Bergeron, C., Rajput, A., Dozic, S., Mastrogiacomo, F., Chang, L.J., Wilson, J.M., DiStefano, L.M., Nobrega, J.N., 1992. Brain cytochrome oxidase in Alzheimer’s disease. J. Neurochem. 59, 776– 779. Kong, J., Xu, Z., 1998. Massive mitochondrial degeneration in motor neurons triggers the onset of amyotrophic lateral sclerosis in mice expressing a mutant SOD1. J. Neurosci. 18 (9), 3241–3250. Korhonen, J.A., Gaspari, M., Falkenberg, M., 2003. TWINKLE Has 50 fi 30 DNA helicase activity and is specifically stimulated by mitochondrial single-stranded DNA-binding protein. J. Biol. Chem. 278 (49), 48627–48632. Kowaltowski, A.J., Smaili, S.S., Russell, J.T., Fiskum, G., 2000. Elevation of resting mitochondrial membrane potential of neural cells by cyclosporin A, BAPTA-AM, and bcl-2. Am. J. Physiol. Cell. Physiol. 279 (3), C852–C859. Lafon-Cazal, M., Pietri, S., Culcasi, M., Bockaert, J., 1993. NMDA-dependent superoxide production and neurotoxicity. Nature 364 (6437), 535–537. Langston, J.W., Ballard Jr., P.A., 1983. Parkinson’s disease in a chemist working with 1-methyl-4-phenyl1,2,5,6-tetrahydropyridine. N. Engl. J. Med. 309 (5), 310. Lasorsa, F.M., Pinton, P., Palmieri, L., Fiermonte, G., Rizzuto, R., Palmieri, F., 2003. Recombinant expression of theCa2þ . -sensitive aspartate/glutamate carrier increases mitochondrial ATP production in agonist-stimulated Chinese hamster ovary cells. J. Biol. Chem. 278 (40), 38686–38692. Lauritzen, I., De Weille, J.R., Lazdunski, M., 1997. The potassium channel opener (-)-cromakalim prevents glutamate-induced cell death in hippocampal neurons. J. Neurochem. 69 (4), 1570–1579. Lawrence, C.L., Billups, B., Rodrigo, G.C., Standen, N.B., 2001. The KATP channel opener diazoxide protects cardiac myocytes during metabolic inhibition without causing mitochondrial depolarization or flavoprotein oxidation. Br. J. Pharmacol. 134 (3), 535–542. Leblanc, C., Richard, O., Kloareg, B., Viehmann, S., Zetsche, K., Boyen, C., 1997. Origin and evolution of mitochondria: what have we learnt from red algae? Curr. Genet. 31, 193–207. Legros, F., Lombes, A., Frachon, P., Rojo, M., 2002. Mitochondrial fusion in human cells is efficient, requires the inner membrane potential, and is mediated by mitofusins. Mol. Biol. Cell. 13, 4343–4354. Lemasters, J.J., Nieminen, A.L., 1999. Negative contrast imaging of mitochondria by confocal microscopy. Biophys. J. 77 (3), 1747–1750. Lesnefsky, E.J., Hoppel, C.L., 2003. Ischemia-reperfusion injury in the aged heart: role of mitochondria. Arch. Biochem. Biophys. 420 (2), 287–297. Leyssens, A., Nowicky, A.V., Patterson, L., Crompton, M., Duchen, M.R., 1996. The relationship between mitochondrial state, ATP hydrolysis, [Mg2þ ]i and [Ca2þ ]i studied in isolated rat cardiomyocytes. J. Physiol. 496 (Pt 1), 111–128. Lin, H., Bhatia, R., Lal, R., 2001. Amyloid b protein forms ion channels: implications for Alzheimer’s disease pathophysiology. FASEB J. 15 (13), 2433–2444. Liu, Y., Sato, T., O’Rourke, B., Marban, E., 1998. Mitochondrial ATP-dependent potassium channels: novel effectors of cardioprotection. Circulation 97 (24), 2463–2469. Loupatatzis, C., Seitz, G., Schonfeld, P., Lang, F., Siemen, D., 2002. Single-channel currents of the permeability transition pore from the inner mitochondrial membrane of rat liver and of a human hepatoma cell line. Cell. Physiol. Biochem. 12 (5–6), 269–278. Lowell, B.B., Spiegelman, B.M., 2000. Towards a molecular understanding of adaptive thermogenesis. Nature 404 (6778), 652–660. Lukyanenko, V., Gyorke, I., Subramanian, S., Smirnov, A., Wiesner, T.F., Gyorke, S., 2000. Inhibition of Ca2þ sparks by ruthenium red in permeabilized rat ventricular myocytes. Biophys. J. 79, 1273– 1284.

446

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

Maechler, P., Wollheim, C.B., 2001. Mitochondrial function in normal and diabetic beta-cells. Nature 414 (6865), 807–812. Mahajan, N.P., Linder, K., Berry, G., Gordon, G.W., Heim, R., Herman, B., 1998. Bcl-2 and Bax interactions in mitochondria probed with green fluorescent protein and fluorescence resonance energy transfer. Nat. Biotechnol. 16, 547–552. Mannella, C.A., Buttle, K., Rath, B.K., Marko, M., 1998. Electron microscopic tomography of rat-liver mitochondria and their interaction with the endoplasmic reticulum. Biofactors (Oxford, England) 8, 225–228. Mannella, C.A., Marko, M., Buttle, K., 1997. Reconsidering mitochondrial structure: new views of an old organelle. Trends Biochem. Sci. 22, 37–38. Mattiasson, G., Shamloo, M., Gido, G., Mathi, K., Tomasevic, G., Yi, S., Warden, C.H., Castilho, R.F., Melcher, T., Gonzalez-Zulueta, M., Nikolich, K., Wieloch, T., 2003. Uncoupling protein-2 prevents neuronal death and diminishes brain dysfunction after stroke and brain trauma. Nat. Med. 9 (8), 1062– 1068. McCormack, J.G., Halestrap, A.P., Denton, R.M., 1990. Role of calcium ions in regulation of mammalian intramitochondrial metabolism. Physiol. Rev. 70 (2), 391–425. Melov, S., 2002a. Therapeutics against mitochondrial oxidative stress in animal models of aging. Ann. NY Acad. Sci. 959, 330–340. Melov, S., 2002b. ‘. . . and C is for Clioquinol’––the AbCs of Alzheimer’s disease. Trends Neurosci. 25, 121–123. Michelakis, E.D., Hampl, V., Nsair, A., Wu, X., Harry, G., Haromy, A., Gurtu, R., Archer, S.L., 2002. Diversity in mitochondrial function explains differences in vascular oxygen sensing. Circ. Res. 90, 1307–1315. Migliaccio, E., Giorgio, M., Mele, S., Pelicci, G., Reboldi, P., Pandolfi, P.P., Lanfrancone, L., Pelicci, P.G., 1999. The p66shc adaptor protein controls oxidative stress response and life span in mammals. Nature 402 (6759), 309–313. Miller, F.J., Rosenfeldt, F.L., Zhang, C., Linnane, A.W., Nagley, P., 2003. Precise determination of mitochondrial DNA copy number in human skeletal and cardiac muscle by a PCR-based assay: lack of change of copy number with age. Nucleic Acids Res. (Online) 31, e61. Mills, E., Jobsis, F.F., 1970. Simultaneous measurement of cytochrome a3 reduction and chemoreceptor afferent activity in the carotid body. Nature 225, 1147–1149. Mills, E., Jobsis, F.F., 1972. Mitochondrial respiratory chain of carotid body and chemoreceptor response to changes in oxygen tension. J. Neurophysiol. 35, 405–428. Miyawaki, A., Griesbeck, O., Heim, R., Tsien, R.Y., 1999. Dynamic and quantitative Ca2þ measurements using improved cameleons. Proc. Natl. Acad. Sci. USA 96, 2135–2140. Mizuno, Y., Ohta, S., Tanaka, M., Takamiya, S., Suzuki, K., Sato, T., Oya, H., Ozawa, T., Kagawa, Y., 1989. Deficiencies in complex I subunits of the respiratory chain in Parkinson’s disease. Biochem. Biophys. Res. Commun. 163 (3), 1450–1455. Momken, I., Fortin, D., Serrurier, B., Bigard, X., Ventura-Clapier, R., Veksler, V., 2002. Endothelial nitric oxide synthase (NOS) deficiency affects energy metabolism pattern in murine oxidative skeletal muscle. Biochem. J. 368 (Pt 1), 341–347. Montero, M., Alonso, M.T., Albillos, A., Garcia-Sancho, J., Alvarez, J., 2001. Mitochondrial Ca2þ induced Ca2þ release mediated by the Ca2þ uniporter. Mol. Biol. Cell. 12, 63–71. Montero, M., Lobaton, C.D., Gutierrez-Fernandez, S., Moreno, A., Alvarez, J., 2003. Modulation of histamine-induced Ca2þ release by protein kinase C. Effects on cytosolic and mitochondrial [Ca2þ ] peaks. J. Biol. Chem. 278 (50), 49972–49979. Montero, M., Lobaton, C.D., Moreno, A., Alvarez, J., 2002. A novel regulatory mechanism of the mitochondrial Ca2þ uniporter revealed by the p38 mitogen-activated protein kinase inhibitor SB202190. FASEB J. 16 (14), 1955–1957. Morais Cardoso, S., Swerdlow, R.H., Oliveira, C.R., 2002. Induction of cytochrome c-mediated apoptosis by amyloid beta 25–35 requires functional mitochondria. Brain Res. 931 (2), 117–125. Muller, W.E., Romero, F.J., Perovic, S., Pergande, G., Pialoglou, P., 1997. Protection of flupirtine on beta-amyloid-induced apoptosis in neuronal cells in vitro: prevention of amyloid-induced glutathione depletion. J. Neurochem. 68 (6), 2371–2377.

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

447

Murata, M., Akao, M., O’Rourke, B., Marban, E., 2001. Mitochondrial ATP-sensitive potassium channels attenuate matrix Ca(2+) overload during simulated ischemia and reperfusion: possible mechanism of cardioprotection. Circ. Res. 89 (10), 891–898. Murphy, T.H., Miyamoto, M., Sastre, A., Schnaar, R.L., Coyle, J.T., 1989. Glutamate toxicity in a neuronal cell line involves inhibition of cystine transport leading to oxidative stress. Neuron 2 (6), 1547–1558. Mutisya, E.M., Bowling, A.C., Beal, M.F., 1994. Cortical cytochrome oxidase activity is reduced in Alzheimer’s disease. J. Neurochem. 63, 2179–2184. Nagai, T., Sawano, A., Park, E.S., Miyawaki, A., 2001. Circularly permuted green fluorescent proteins engineered to sense Ca2þ . Proc. Natl. Acad. Sci. USA 98 (6), 3197–3202. Nakada, K., Inoue, K., Ono, T., Isobe, K., Ogura, A., Goto, Y.I., Nonaka, I., Hayashi, J.I., 2001. Intermitochondrial complementation: Mitochondria-specific system preventing mice from expression of disease phenotypes by mutant mtDNA. Nat. Med. 7, 934–940. N€ aslund, J., Haroutunian, V., Mohs, R., Davis, K.-L., Davies, P., Greengard, P., Buxbaum, J.-D., 2000. Correlation between elevated levels of amyloid b-peptide in the brain and cognitive decline. JAMA 283, 1571–1577. Neupert, W., 1997. Protein import into mitochondria. Ann. Rev. Biochem. 66, 863–917. Nicholls, D.G., 2002. Mitochondrial function and dysfunction in the cell: its relevance to aging and agingrelated disease. Int. J. Biochem. Cell Biol. 34 (11), 1372–1381. Nicholls, D.G., Budd, S.L., 2000. Mitochondria and neuronal survival. Physiol. Rev. 80 (1), 315–360. Nicholls, D.G., Crompton, M., 1980. Mitochondrial calcium transport. FEBS Lett. 111, 261–268. Nieminen, A.L., Saylor, A.K., Tesfai, S.A., Herman, B., Lemasters, J.J., 1995. Contribution of the mitochondrial permeability transition to lethal injury after exposure of hepatocytes to t-butylhydroperoxide. Biochem. J. 307 (Pt 1), 99–106. Nisoli, E., Clementi, E., Paolucci, C., Cozzi, V., Tonello, C., Sciorati, C., Bracale, R., Valerio, A., Francolini, M., Moncada, S., Carruba, M.O., 2003. Mitochondrial biogenesis in mammals: the role of endogenous nitric oxide. Science 299 (5608), 896–899. Nitschke, R., Wilhelm, S., Borlinghaus, R., Leipziger, J., Bindels, R., Greger, R., 1997. A modified confocal laser scanning microscope allows fast ratio imaging of intracellular Ca2þ activity using Fura2. Pflug. Arch.: Eur. J. Physiol. 433, 653–663. O’Reilly, C.M., Fogarty, K.E., Drummond, R.M., Tuft, R.A., Walsh Jr., J.V., 2004. Spontaneous mitochondrial depolarizations are independent of SR Ca2þ release. Am. J. Physiol. Cell. Physiol. 286 (5), C1139–C1151. Ortega-Saenz, P., Pardal, R., Garcia-Fernandez, M., Lopez-Barneo, J., 2003. Rotenone selectively occludes sensitivity to hypoxia in rat carotid body glomus cells. J. Physiol. 548 (Pt 3), 789– 800. Ovide-Bordeaux, S., Ventura-Clapier, R., Veksler, V., 2000. Do modulators of the mitochondrial K(ATP) channel change the function of mitochondria in situ? J. Biol. Chem. 275 (47), 37291–37295. Pacher, P., Hajnoczky, G., 2001. Propagation of the apoptotic signal by mitochondrial waves. EMBO J. 20 (15), 4107–4121. Pacher, P., Thomas, A.P., Hajnoczky, G., 2002. Ca2þ marks: miniature calcium signals in single mitochondria driven by ryanodine receptors. Proc. Natl. Acad. Sci. USA 99 (4), 2380–2385. Palecek, J., Lips, M.B., Keller, B.U., 1999. Calcium dynamics and buffering in motoneurones of the mouse spinal cord. J. Physiol. 520 (Pt 2), 485–502. Panov, A.V., Gutekunst, C.A., Leavitt, B.R., Hayden, M.R., Burke, J.R., Strittmatter, W.J., Greenamyre, J.T., 2002. Early mitochondrial calcium defects in Huntington’s disease are a direct effect of polyglutamines. Nat. Neurosci. 5 (8), 731–736. Park, M.K., Ashby, M.C., Erdemli, G., Petersen, O.H., Tepikin, A.V., 2001. Perinuclear, perigranular and sub- lasmalemmal mitochondria have distinct functions in the regulation of cellular calcium transport. EMBO J. 20, 1863–1874. Parker Jr., W.D., Boyson, S.J., Parks, J.K., 1989. Abnormalities of the electron transport chain in idiopathic Parkinson’s disease. Ann. Neurol. 26 (6), 719–723. Parpura-Gill, A., Beitz, D., Uemura E, 1997. The inhibitory effects of beta-amyloid on glutamate and glucose uptakes by cultured astrocytes. Brain Res. 754 (1–2), 65–71.

448

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

Pastorino, J.G., Tafani, M., Rothman, R.J., Marcinkeviciute, A., Hoek, J.B., Farber, J.L., Marcineviciute, A., 1999. Functional consequences of the sustained or transient activation by Bax of the mitochondrial permeability transition pore. J. Biol. Chem. 274 (44), 31734–31739. Peachman, K.K., Lyles, D.S., Bass, D.A., 2001. Mitochondria in eosinophils: functional role in apoptosis but not respiration. Proc. Natl. Acad. Sci. USA 98 (4), 1717–1722. Peng, T.I., Greenamyre, J.T., 1998. Privileged access to mitochondria of calcium influx through N-methylD-aspartate receptors. Mol. Pharmacol. 53 (6), 974–980. Perkins, G., Renken, C., Martone, M.E., Young, S.J., Ellisman, M., Frey, T., 1997. Electron tomography of neuronal mitochondria: three-dimensional structure and organization of cristae and membrane contacts. J. Struct. Biol. 119, 260–272. Perkins, G.A., Frey, T.G., 2000. Recent structural insight into mitochondria gained by microscopy. Micron (Oxford, England: 1993) 31, 97–111. Perkins, G.A., Renken, C.W., Frey, T.G., Ellisman, M.H., 2001. Membrane architecture of mitochondria in neurons of the central nervous system. J. Neurosci. Res. 66, 857–865. Perkins, G.A., Song, J.Y., Tarsa, L., Deerinck, T.J., Ellisman, M.H., Frey, T.G., 1998. Electron tomography of mitochondria from brown adipocytes reveals crista junctions. J. Bioenerg. Biomembr. 30, 431–442. Petronilli, V., Miotto, G., Canton, M., Brini, M., Colonna, R., Bernardi, P., Di Lisa, F., 1999. Transient and long-lasting openings of the mitochondrial permeability transition pore can be monitored directly in intact cells by changes in mitochondrial calcein fluorescence. Biophys. J. 76 (2), 725–734. Petronilli, V., Miotto, G., Canton, M., Colonna, R., Bernardi, P., Di Lisa, F., 1998. Imaging the mitochondrial permeability transition pore in intact cells. Biofactors 8 (3–4), 263–272. Petronilli, V., Penzo, D., Scorrano, L., Bernardi, P., Di Lisa, F., 2001. The mitochondrial permeability transition, release of cytochrome c and cell death. Correlation with the duration of pore openings in situ. J. Biol. Chem. 276 (15), 12030–12034. Pfanner, N., Truscott, K.N., 2002. Powering mitochondrial protein import. Nat. Struct. Biol. 9 (4), 234– 236. Pralong, W.F., Hunyady, L., Varnai, P., Wollheim, C.B., Spat, A., 1992. Pyridine nucleotide redox state parallels production of aldosterone in potassium-stimulated adrenal glomerulosa cells. Proc. Natl. Acad. Sci. USA 89 (1), 132–136. Rakhit, R.D., Mojet, M.H., Marber, M.S., Duchen, M.R., 2001. Mitochondria as targets for nitric oxideinduced protection during simulated ischemia and reoxygenation in isolated neonatal cardiomyocytes. Circulation 103 (21), 2617–2623. Rao, S.D., Yin, H.Z., Weiss, J.H., 2003. Disruption of glial glutamate transport by reactive oxygen species produced in motor neurons. J. Neurosci. 23 (7), 2627–2633. Renken, C., Siragusa, G., Perkins, G., Washington, L., Nulton, J., Salamon, P., Frey, T.G., 2002. A thermodynamic model describing the nature of the crista junction: a structural motif in the mitochondrion. J. Struct. Biol. 138, 137–144. Reshef, A., Sperling, O., Zoref-Shani, E., 1998. Opening of ATP-sensitive potassium channels by cromakalim confers tolerance against chemical ischemia in rat neuronal cultures. Neurosci. Lett. 250 (2), 111–114. Reynier, P., May_Panloup, P., Chretien, M.F., Morgan, C.J., Jean, M., Savagner, F., Barriere, P., Malthiery, Y., 2001. Mitochondrial DNA content affects the fertilizability of human oocytes. Mol. Human Reprod. 7, 425–429. Rhee, S.K., Quist, A.P., Lal, R., 1998. Amyloid b protein-(1-42) forms calcium-permeable, Zn2þ -sensitive channel. J. Biol. Chem. 273, 13379–13382. Rizzuto, R., Bastianutto, C., Brini, M., Murgia, M., Pozzan, T., 1994. Mitochondrial Ca2þ homeostasis in intact cells. J. Cell Biol. 126, 1183–1194. Rizzuto, R., Brini, M., Murgia, M., Pozzan, T., 1993. Microdomains with high Ca2þ close to IP3-sensitive channels that are sensed by neighboring mitochondria. Science 262, 744–747. Rizzuto, R., Duchen, M.R., Pozzan, T., 2004. Flirting in little space: the ER/mitochondria Ca2þ liaison. Sci. STKE 2004 (215). Rizzuto, R., Simpson, A.W., Brini, M., Pozzan, T., 1992. Ca2þ revealed by specifically targeted recombinant aequorin. Rapid changes of mitochondrial. Nature 358, 325–327.

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

449

Robb-Gaspers, L.D., Burnett, P., Rutter, G.A., Denton, R.M., Rizzuto, R., Thomas, A.P., 1998. Integrating cytosolic calcium signals into mitochondrial metabolic responses. EMBO J. 17, 4987–5000. Rodrigo, G.C., Lawrence, C.L., Standen, N.B., 2002. Dinitrophenol pretreatment of rat ventricular myocytes protects against damage by metabolic inhibition and reperfusion. J. Mol. Cell. Cardiol. 34 (5), 555–569. Rottenberg, H., Wu, S., 1998. Quantitative assay by flow cytometry of the mitochondrial membrane potential in intact cells. Biochim. Biophys. Acta 1404, 393–404. Rouslin, W., 1991. Regulation of the mitochondrial ATPase in situ in cardiac muscle: role of the inhibitor subunit. J. Bioenerg. Biomembr. 23 (6), 873–888. Rovira, C., Arbez, N., Mariani, J., 2002. Ab (25-35) and Ab (1-40) act on different calcium channels in CA1 hippocampal neurons. Biochem. Biophys. Res. Commun. 296, 1317–1321. Rutter, G.A., Rizzuto, R., 2000. Regulation of mitochondrial metabolism by ER Ca2þ release: an intimate connection. Trends Biochem. Sci. 25 (5), 215–221. Sattler, R., Tymianski, M., 2001. Molecular mechanisms of glutamate receptor-mediated excitotoxic neuronal cell death. Mol. Neurobiol. 24 (1–3), 107–129. Sattler, R., Xiong, Z., Lu, W.Y., Hafner, M., MacDonald, J.F., Tymianski, M., 1999. Specific coupling of NMDA receptor activation to nitric oxide neurotoxicity by PSD-95 protein. Science 284 (5421), 1845– 1848. Schapira, A.H., Cooper, J.M., Dexter, D., Jenner, P., Clark, J.B., Marsden, C.D., 1989. Mitochondrial complex I deficiency in Parkinson’s disease. Lancet 1 (8649), 1269. Schubert, D., Piasecki, D., 2001. Oxidative glutamate toxicity can be a component of the excitotoxicity cascade. J. Neurosci. 21 (19), 7455–7462. Schwartz, M., Vissing, J., 2002. Paternal inheritance of mitochondrial DNA. New Engl. J. Med. 347, 576– 580. Sharma, V.K., Ramesh, V., Franzini-Armstrong, C., Sheu, S.S., 2000. Transport of Ca2þ from sarcoplasmic reticulum to mitochondria in rat ventricular myocytes. J. Bioenerg. Biomembr. 32 (1), 97–104. Sheehan, J.P., Swerdlow, R.H., Miller, S.W., Davis, R.E., Parks, J.K., Parker, W.D., Tuttle, J.B., 1997. Calcium homeostasis and reactive oxygen species production in cells transformed by mitochondria from individuals with sporadic Alzheimer’s disease. J. Neurosci. 17 (12), 4612–4622. Sherer, T.B., Betarbet, R., Stout, A.K., Lund, S., Baptista, M., Panov, A.V., Cookson, M.R., Greenamyre, J.T., 2002. An in vitro model of Parkinson’s disease: linking mitochondrial impairment to altered alpha-synuclein metabolism and oxidative damage. J. Neurosci. 22 (16), 7006–70015. Sherer, T.B., Betarbet, R., Testa, C.M., Seo, B.B., Richardson, J.R., Kim, J.H., Miller, G.W., Yagi, T., Matsuno-Yagi, A., Greenamyre, J.T., 2003. Mechanism of toxicity in rotenone models of Parkinson’s disease. J. Neurosci. 23 (34), 10756–10764. Sheu, K.F., Kim, Y.T., Blass, J.P., Weksler, M.E., 1985. An immunochemical study of the pyruvate dehydrogenase deficit in Alzheimer’s disease brain. Ann. Neurol. 17, 444–449. Shimizu, S., Narita, M., Tsujimoto, Y., 1999. Bcl-2 family proteins regulate the release of apoptogenic cytochrome c by the mitochondrial channel VDAC. Nature, 399(6735):483-7. Erratum in: Nature 2000 Oct 12;407(6805):767. Siemering, K.R., Golbik, R., Sever, R., Haseloff, J., 1996. Mutations that suppress the thermosensitivity of green fluorescent protein. Curr. Biol. 6, 1653–1663. Skulachev, V.P., 2000. Mitochondria in the programmed death phenomena; a principle of biology: ‘‘it is better to die than to be wrong’’. IUBMB Life 49 (5), 365–373. Smith, M.A., Richey Harris, P.L., Sayre, L.M., Beckman, J.S., Perry, G., 1997. Widespread peroxynitritemediated damage in Alzheimer’s disease. J. Neurosci. 17, 2653–2657. Snowdon, D.A., 1997. Aging and Alzheimer’s disease: lessons from the Nun Study. Gerontologist 37, 150– 156. Sparagna, G.C., Gunter, K.K., Sheu, S.S., Gunter, T.E., 1995. Mitochondrial calcium uptake from physiological-type pulses of calcium. A description of the rapid uptake mode. J. Biol. Chem. 270, 27510–27515. Spelbrink, J.N., Li, F.Y., Tiranti, V., Nikali, K., Yuan, Q.P., Tariq, M., Wanrooij, S., Garrido, N., Comi, G., Morandi, L., Santoro, L., Toscano, A., Fabrizi, G.M., Somer, H., Croxen, R., Beeson, D.,

450

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

Poulton, J., Suomalainen, A., Jacobs, H.T., Zeviani, M., Larsson, C., 2001. Human mitochondrial DNA deletions associated with mutations in the gene encoding Twinkle, a phage T7 gene 4-like protein localized in mitochondria. Nat. Genet. 28 (3), 223–231. Starkov, A.A., Polster, B.M., Fiskum, G., 2002. Regulation of hydrogen peroxide production by brain mitochondria by calcium and Bax. J. Neurochem. 83 (1), 220–228. Streller, T., Huckstorf, C., Pfeiffer, C., Acker, H., 2002. Unusual cytochrome a592 with low PO2 affinity correlates as putative oxygen sensor with rat carotid body chemoreceptor discharge. Faseb J. 16, 1277– 1279. Stock, D., Leslie, A.G., Walker, J.E., 1999. Molecular architecture of the rotary motor in ATP synthase. Science 286, 1700–1705. Stout, A.K., Raphael, H.M., Kanterewicz, B.I., Klann, E., Reynolds, I.J., 1998. Glutamate-induced neuron death requires mitochondrial calcium uptake. Nat. Neurosci. 1 (5), 366–373. St-Pierre, J., Buckingham, J.A., Roebuck, S.J., Brand, M.D., 2002. Topology of superoxide production from different sites in the mitochondrial electron transport chain. J. Biol. Chem. 277 (47), 44784–44790. Sutovsky, P., Moreno, R.D., Ramalho_Santos, J., Dominko, T., Simerly, C., Schatten, G., 2000. Ubiquitinated sperm mitochondria, selective proteolysis, and the regulation of mitochondrial inheritance in mammalian embryos. Biol. Reprod. 63, 582–590. Swerdlow, R.H., Parks, J.K., Keeney, P., Bennett Jr., J.P., Miller, S.W., Davis, R.E., Parker Jr., W.D., 2000. Abnormal mitochondrial morphology in sporadic Parkinson’s and Alzheimer’s disease cybrid cell lines. Exp. Neurol. 162, 37–50. Szabadkai, G., Pitter, J.G., Spat, A., 2001. Cytoplasmic Ca2þ at low submicromolar concentration stimulates mitochondrial metabolism in rat luteal cells. Pflugers Arch. 441 (5), 678–685. Szalai, G., Krishnamurthy, R., Hajnoczky, G., 1999. Apoptosis driven by IP(3)-linked mitochondrial calcium signals. EMBO J. 18 (22), 6349–63361. Szatkowski, M., Attwell, D., 1994. Triggering and execution of neuronal death in brain ischaemia: two phases of glutamate release by different mechanisms. Trends Neurosci. 17 (9), 359–365. Tang, Y., Zucker, R.S., 1997. Mitochondrial involvement in post-tetanic potentiation of synaptic transmission. Neuron 18, 483–491. Thayer, S.A., Miller, R.J., 1990. Regulation of the intracellular free calcium concentration in single rat dorsal root ganglion neurones in vitro. J. Physiol. 425, 85–115. Tinel, H., Cancela, J.M., Mogami, H., Gerasimenko, J.V., Gerasimenko, O.V., Tepikin, A.V., Petersen, O.H., 1999. Active mitochondria surrounding the pancreatic acinar granule region prevent spreading of inositol trisphosphate-evoked local cytosolic Ca2þ signals. EMBO J. 18, 4999–5008. Tretter, L., Adam-Vizi, V., 2000. Inhibition of Krebs cycle enzymes by hydrogen peroxide: A key role of aketoglutarate dehydrogenase in limiting NADH production under oxidative stress. J. Neurosci. 20, 8972–8979. Tymianski, M., Charlton, M.P., Carlen, P.L., Tator, C.H., 1993. Source specificity of early calcium neurotoxicity in cultured embryonic spinal neurons. J. Neurosci. 13 (5), 2085–2104. Van Den Bosch, L., Schwaller, B., Vleminckx, V., Meijers, B., Stork, S., Ruehlicke, T., Van Houtte, E., Klaassen, H., Celio, M.R., Missiaen, L., Robberecht, W., Berchtold, M.W., 2002. Protective effect of parvalbumin on excitotoxic motor neuron death. Exp. Neurol. 174 (2), 150–161. van-der-Leij, F.R., Kram, A.M., Bartelds, B., Roelofsen, H., Smid, G.B., Takens, J., Zammit, V.A., Kuipers, J.R., 1999. Cytological evidence that the C-terminus of carnitine palmitoyltransferase I is on the cytosolic face of the mitochondrial outer membrane. Biochem. J. 341 (Pt 3), 777–784. Varadarajan, S., Yatin, S., Aksenova, M., Butterfield, D.A., 2000. Alzheimer’s amyloid b-peptideassociated free radical oxidative stress and neurotoxicity. J. Struct. Biol. 30, 184–208. Vergun, O., Keelan, J., Khodorov, B.I., Duchen, M.R., 1999. Glutamate-induced mitochondrial depolarisation and perturbation of calcium homeostasis in cultured rat hippocampal neurones. J. Physiol. 519 (Pt 2), 451–466. Vergun, O., Sobolevsky, A.I., Yelshansky, M.V., Keelan, J., Khodorov, B.I., Duchen, M.R., 2001. Exploration of the role of reactive oxygen species in glutamate neurotoxicity in rat hippocampal neurones in culture. J. Physiol. 531 (Pt 1), 147–163. Vergun, O., Votyakova, T.V., Reynolds, I.J., 2003. Spontaneous changes in mitochondrial membrane potential in single isolated brain mitochondria. Biophys J. 85 (5), 3358–3366.

M.R. Duchen / Molecular Aspects of Medicine 25 (2004) 365–451

451

Voos, W., 2003. A new connection: chaperones meet a mitochondrial receptor. Mol. Cell. 11 (1), 1–3. Votyakova, T.V., Reynolds, I.J., 2001. DeltaPsi(m)-Dependent and -independent production of reactive oxygen species by rat brain mitochondria. J. Neurochem. 79 (2), 266–277. Wang, S., Hazelrigg, T., 1994. Implications for bcd mRNA localization from spatial distribution of exu protein in Drosophila oogenesis. Nature 369, 400–403. Ward, J.P., 2003. Mitochondria and oxygen sensing: fueling the controversy. J. Physiol. 548, 664. Ward, J.P., Aaronson, P.I., 1999. Mechanisms of hypoxic pulmonary vasoconstriction: can anyone be right? Respirat. Physiol. 115, 261–271. Ward, M.W., Rego, A.C., Frenguelli, B.G., Nicholls, D.G., 2000. Mitochondrial membrane potential and glutamate excitotoxicity in cultured cerebellar granule cells. J. Neurosci. 20, 7208–7219. Williams, B.A., Buckler, K.J., 2000. Identification of an oxygen-sensitive potassium channel in neonatal rat carotid body type I cells. Adv. Exp. Med. Biol. 475, 419–424. Wong, E.D., Wagner, J.A., Gorsich, S.W., McCaffery, J.M., Shaw, J.M., Nunnari, J., 2000. The dynaminrelated GTPase, Mgm1p, is an intermembrane space protein required for maintenance of fusion competent mitochondria. J. Cell Biol. 151 (2), 341–352. Wood, J.N., Winter, J., James, I.F., Rang, H.P., Yeats, J., Bevan, S., 1998. Capsaicin-induced ion fluxes in dorsal root ganglion cells in culture. J. Neurosci. 8, 3208–3220. Yellon, D.M., Downey, J.M., 2003. Preconditioning the myocardium: from cellular physiology to clinical cardiology. Physiol. Rev. 83 (4), 1113–1151. Zamzami, N., Susin, S.A., Marchetti, P., Hirsch, T., Gomez-Monterrey, I., Castedo, M., Kroemer, G., 1996. Mitochondrial control of nuclear apoptosis. J. Exp. Med. 183 (4), 1533–1544. Zimmermann, B., 2000. Control of InsP3-induced Ca2þ oscillations in permeabilized blowfly salivary gland cells: contribution of mitochondria. J. Physiol. 525, 707–719. Zoratti, M., Szabo, I., 1995. The mitochondrial permeability transition. Biochim. Biophys. Acta 1241 (2), 139–176.