OUP user menu

Mitochondrial permeability transition pore opening during myocardial reperfusion—a target for cardioprotection

Andrew P Halestrap, Samantha J Clarke, Sabzali A Javadov
DOI: http://dx.doi.org/10.1016/S0008-6363(03)00533-9 372-385 First published online: 15 February 2004


Reperfusion of the heart after a period of ischaemia leads to the opening of a nonspecific pore in the inner mitochondrial membrane, known as the mitochondrial permeability transition pore (MPTP). This transition causes mitochondria to become uncoupled and capable of hydrolysing rather than synthesising ATP. Unrestrained, this will lead to the loss of ionic homeostasis and ultimately necrotic cell death. The functional recovery of the Langendorff-perfused heart from ischaemia inversely correlates with the extent of pore opening, and inhibition of the MPTP provides protection against reperfusion injury. This may be mediated either by a direct interaction with the MPTP [e.g., by Cyclosporin A (CsA) and Sanglifehrin A (SfA)], or indirectly by decreasing calcium loading and reactive oxygen species (ROS; key inducers of pore opening) or lowering intracellular pH. Agents working in this way may include pyruvate, propofol, Na+/H+ antiporter inhibitors, and ischaemic preconditioning (IPC). Mitochondrial KATP channels have been implicated in preconditioning, but our own data suggest that the channel openers and blockers used in these studies work through alternative mechanisms. In addition to its role in necrosis, transient opening of the MPTP may occur and lead to the release of cytochrome c and other proapoptotic molecules that initiate the apoptotic cascade. However, only if subsequent MPTP closure occurs will ATP levels be maintained, ensuring that cell death continues down an apoptotic, rather than a necrotic, pathway.

  • Ischaemic preconditioning
  • Mitochondrial potassium channels
  • Calcium overload
  • Reactive oxygen species
  • Apoptosis

1 Introduction

Mitochondria play critical roles in both the life and death of cells. In healthy cardiac myocytes, their primary function is the provision of ATP through oxidative phosphorylation to meet the high energy demands of the beating heart. Glycolysis alone is unable to meet these demands even in the resting state and inhibition of oxidative phosphorylation, as occurs in anoxia or ischaemia, leads to impairment or cessation of normal heart function. However, latent within mitochondria, there exist mechanisms that, once activated, convert the mitochondria from organelles that support the life of the cell to those that actively induce both apoptotic and necrotic cell death. The switch in roles, akin to the conversion of Dr. Jekyll to Mr. Hyde, is mediated by the opening of a nonspecific pore in the mitochondrial inner membrane, known as the mitochondrial permeability transition pore (MPTP). This normally remains closed, but can open under conditions of cellular stress with dire consequences. In this review, we will briefly summarise what is known about the molecular mechanism of the MPTP and why it opens during reperfusion. We will then describe the techniques that have been used to measure pore opening in the perfused heart, and how these experiments have led to the proposal that the extent of pore opening is a critical determinant of reperfusion injury. Finally, we will describe how inhibiting MPTP opening is an effective strategy for the protection of hearts from reperfusion injury.

2 The mitochondrial permeability transition pore

2.1 Causes and consequences of MPTP opening

Under normal physiological conditions, the mitochondrial inner membrane is impermeable to all, but a few, selected metabolites and ions. However, under conditions of stress, a nonspecific pore known as the mitochondrial permeability transition pore can open in the mitochondrial inner membrane that allows free passage of any molecule of <1.5 kDa [1–3]. When the MPTP opens, the permeability barrier of the inner membrane becomes disrupted with two major consequences. First, although all small molecular weight solutes move freely across the membrane, proteins do not and, as a result, they exert a colloidal osmotic pressure that causes mitochondria to swell. Although the unfolding of the cristae allows the matrix to expand without rupture of the inner membrane, the outer membrane will break and lead to the release of proteins in the intermembrane space such as cytochrome c and other factors that play a critical role in apoptotic cell death (see Section 3.2.1). Second, the inner membrane becomes freely permeable to protons. This uncouples oxidative phosphorylation, causing the proton-translocating ATPase to reverse direction and so actively hydrolyse ATP, rather than synthesis it. Under such conditions, intracellular ATP concentrations rapidly decline, leading to the disruption of ionic and metabolic homeostasis and the activation of degradative enzymes such as phospholipases, nucleases, and proteases. Unless pore closure occurs, these changes will cause irreversible damage to the cell, resulting in necrotic death. Even if closure does occur, the mitochondrial swelling and outer membrane rupture may be sufficient to set the apoptotic cascade in motion. Thus, it is hardly surprising that the MPTP is kept firmly closed under normal physiological conditions and is only activated under pathological conditions. The key factor responsible for MPTP opening is mitochondrial calcium overload (i.e., when mitochondrial matrix [Ca2+] is greatly increased), especially when this is accompanied by oxidative stress, adenine nucleotide depletion, elevated phosphate concentrations, and mitochondrial depolarisation [1,4]. These conditions are exactly those that the heart experiences during postischaemic reperfusion, and there is increasing evidence that MPTP opening is critical in the transition from reversible to irreversible reperfusion injury [1,4].

2.2 The molecular mechanism of the MPTP

For a detailed account of the molecular mechanism of the MPTP, including the experimental evidence on which this is based, the reader is directed to other recent reviews from this and other laboratories [1,3,5]. As illustrated in Fig. 1, the core components of the MPTP are the adenine nucleotide translocase (ANT) and a mitochondrial cyclophilin D (CyP-D) that exhibits peptidyl-prolyl cistrans isomerase (PPIase) activity. Triggered by Ca2+, whose binding to the ANT is inhibited by adenine nucleotides, this PPIase activity causes a conformational change of the carrier that converts it into a nonspecific pore. Cyclosporin A (CsA) acts as a potent inhibitor of MPTP opening (K0.5=5 nM) by preventing CyP-D binding to the ANT. Very recently, we have described another extremely potent inhibitor of the MPTP, Sanglifehrin A (SfA), which is unrelated to CsA [6]. SfA does not prevent CyP-D binding to the ANT but does inhibit its PPIase activity (K0.5<5 nM), preventing it from facilitating the conformational change of the ANT required for pore formation. An important advantage of SfA is that, unlike the complex of CsA with cytosolic cyclophilin A (CyP-A), the SfA–CyP-A complex has no effect on the calcium-activated protein phosphatase, calcineurin. Inhibition of the MPTP by both SfA and CsA can be overcome at high [Ca2+], suggesting that the conformational change facilitated by CyP-D can occur in its absence. Indeed, reconstitution studies have shown that at high [Ca2+], the purified ANT can undergo a conformational change to produce a nonspecific channel in the absence of CyP-D, although the sensitivity to [Ca2+] is enhanced by its presence [7,8]. Other proteins have been implicated in MPTP formation by many workers including the outer membrane voltage-dependent anion channel (VDAC; also known as porin) and the peripheral benzodiazepine receptor [1,3,5]. Although there is debate over whether such proteins are essential structural components or merely exert an important regulatory role, it does seem probable that the MPTP is associated with contact sites between the outer and inner mitochondrial membranes where VDAC and the ANT are thought to interact [1,3,5].

Fig. 1

The proposed molecular mechanism of the mitochondrial permeability transition pore. The probable sites of action of known effectors of pore opening are shown in Table 1.

Numerous factors can influence the sensitivity of MPTP opening to [Ca2+] and these are summarised in Table 1. Any intervention that reduces adenine nucleotide binding to the ANT enhances pore opening, including adenine nucleotide depletion, matrix phosphate (competes for the nucleotide binding site), and the conformational state of the ANT. The latter can be influenced by specific ligands of the ANT such as carboxyatractyloside (CAT; decreases matrix ADP binding affinity) and bongkrekic acid (BKA; increases matrix ADP binding affinity) as well as by the membrane potential (Δψ). Of particular relevance to reperfusion injury, we have shown that oxidative stress also sensitises the MPTP to [Ca2+] by antagonising adenine nucleotide binding, but in addition, it increases CyP-D binding to the ANT. Low pH (<7) and Mg2+ inhibit MPTP opening by antagonising Ca2+ binding [3,5]. The modes of action of two other potent inhibitors of the MPTP are not clear. Trifluoperazine is a potent inhibitor of the MPTP under energised, but not deenergised, conditions and may act by altering surface membrane charge and hence the voltage sensitivity of the MPTP [3]. More recently, it has been shown that ubiquinone analogues can act either as potent activators or inhibitors of the MPTP perhaps by interacting with Complex 1 of the respiratory chain, although how this might regulate the MPTP remains uncertain [9].

View this table:
Table 1

Proposed sites of action of known effectors of the mitochondrial permeability transition

Effect via change in CyP-D binding to the ANTaEffect via change in nucleotide binding to the ANTaDirect effect on Ca2+ binding to the ANTUnknown mode of action
Oxidative stress (e.g., reperfusion, t-butyl hydroperoxide, or diamide) or vicinal thiol reagents (e.g., phenylarsine oxide) to cross-link Cys159 with Cys256 of the ANTThiol reagents attacking Cys159 of ANT (e.g., eosine maleimide)High pHSome ubiquinone analogues (e.g., decyl-ubiquinone, ubiquinone 10)
Increased matrix volumeOxidative stress (e.g., reperfusion, t-butyl hydroperoxide, or diamide) or thiol reagents (e.g., phenylarsine oxide and eosine maleimide) that modify Cys159 of the ANT
Chaotropic agents“c” Conformation of ANT as induced by carboxyatractyloside
Adenine nucleotide depletion
High matrix [Pi] and [PPi]
CsA and some analogues (e.g., Cyclosporin G, 6-methyl-ala-CsA, and 4-methyl-val-CsA)Increase membrane potentialLow pHSome ubiquinone analogues (e.g., ubiquinone 0, 2,5-dihydroxy-6-undecyl-1,4-benzoquinone)
SfA (inhibits PPIase activity of CyP-D but not binding)“m” Conformation of ANT as induced by bongkrekic acidMg2+, Mn2+, Sr2+, Ba2+Trifluoperazine (may work via membrane surface charge)
  • Further details may be found in the text and in Refs. [3–5].

  • a Note that both CyP-D binding and ADP binding exert their effects through changes in the sensitivity of the MPT to [Ca2+].

3 The mitochondrial permeability transition pore opens during reperfusion but not ischaemia

3.1 The conditions prevailing during ischaemia and reperfusion favour pore opening

As reviewed elsewhere [4] and summarised in Fig. 2, the conditions that occur during reperfusion are exactly those that induce MPTP opening. In outline, the increase in glycolysis that occurs during myocardial ischaemia causes a progressive accumulation of lactic acid and drop in pHi that will eventually inhibit further glycolytic flux and ATP production. Activation of the Na+/H+ antiporter occurs as the cell attempts to restore the pHi,, but in the process, it becomes loaded with sodium, which cannot be pumped out of the cell because the Na/K ATPase is inhibited by the lack of ATP. Consequently, the activity of the Na+/Ca2+ antiporter, which usually pumps Ca2+ out of the cell, is reduced or reversed and the cell becomes loaded with calcium. At this point, ionic homeostasis can no longer be maintained, intracellular [Na+] and [Ca2+] begin to rise, and contracture follows (reviewed in Ref. [10]). In addition, adenine nucleotides are degraded during ischaemia and the resulting decrease in adenine nucleotide concentration, in conjunction with the increased phosphate concentrations, will sensitise MPTP opening to [Ca2+] as outlined above. However, because the pHi remains low, the MPTP will not actually open, although if the period of ischaemia is sufficiently prolonged, the heart will become irreversibly damaged through the action of degradative enzymes such as phospholipases and proteases, which themselves will compromise mitochondrial function.

Fig. 2

Factors leading to opening of the MPTP in reperfusion. Further details are given in the text.

In order to salvage the ischaemic heart, it must be reperfused, yet the very process of reperfusion may exacerbate the damage induced by ischaemia itself, as reflected in changes in cell morphology typical of necrosis and the release of intracellular enzymes. Another feature is the appearance of swollen dysfunctional mitochondria that are typical of those that have undergone the permeability transition, and there are good theoretical reasons why this should happen. Upon reperfusion, the mitochondria are once again able to respire and generate a membrane potential to drive ATP synthesis, but this also enables the rapid accumulation of calcium within the mitochondria, leading to calcium overload (see Ref. [10]). In addition, there is a rapid and extensive production of reactive oxygen species (ROS) as the inhibited respiratory chain is reexposed to oxygen. Conditions are now almost optimal for MPTP formation since there is high intramitochondrial [Ca2+] and phosphate, oxidative stress, and depleted adenine nucleotide concentrations. The one restraining influence is the low pHi, but within a few minutes of reperfusion, this returns to normal values and all restraints on MPTP opening are removed.

3.2 Experimental demonstration that the MPTP opens during reperfusion

In order to demonstrate directly that the MPTP only opens during reperfusion and not during ischaemia, three distinct approaches have been taken. First, fluorescence microscopy of isolated cardiac myocytes subjected to simulated ischaemia and reperfusion has been used. In these studies, the mitochondrial membrane potential was determined with fluorescent dyes as a surrogate indicator of MPTP opening [11]. Such dyes accumulate within the mitochondria in response to the membrane potential and are released upon depolarisation when the MPTP opens. Confirmation that any depolarisation is caused by the opening of the MPTP is usually provided by the ability of CsA, sometimes supplemented with trifluoperazine, to inhibit the process [12]. A more sophisticated approach is to determine simultaneously the distribution of another fluorescent dye, calcein (green fluorescence), that can only cross the inner mitochondrial membrane when the pore opens [13,14]. Although this technique has been applied successfully to cardiac myocytes subject to oxidative stress [15–17], the use of isolated cardiac myocytes cannot accurately mimic the situation in the ischaemic/reperfused heart.

Two techniques have been developed to determine the extent of MPTP opening in the perfused heart. One, devised by DiLisa et al. [18], is to determine the loss of mitochondrial NAD+ that accompanies reperfusion as a surrogate indicator of pore opening. Both mitochondrial and cytosolic NAD+ are lost during reperfusion. The latter correlates with the extent of lactate dehydrogenase release as might be predicted since this reflects a breakdown of the plasma membrane permeability barrier. The loss of NAD+ is inhibited by CsA, confirming that it occurs as a result of MPTP opening, although it is impossible to rule out that some of the CsA-sensitive pore openings and disruption of mitochondria detected by this technique actually occur during mitochondrial isolation rather than in situ [6,19]. Another approach, devised in this laboratory, circumvents this potential problem by measuring the mitochondrial entrapment of a radioactive marker, which has been loaded into the cytosol of the heart.

In this technique, illustrated in Fig. 3, Langendorff-perfused hearts are first loaded with [3H]2-deoxyglucose (3H-DOG), which accumulates within the cytosol as 3H-DOG-6-phosphate (3H-DOG-P) but can only enter the mitochondria when the MPTP opens [19]. The extent to which the 3H-DOG-P enters the mitochondria can be determined by their rapid isolation in the presence of EGTA, which reseals the pores and so entraps the 3H-DOG-P within the mitochondria. Measurement of the 3H content of the mitochondria gives a quantitative value for the extent of pore opening, provided that suitable controls and corrections are performed to account for variations in the initial loading of the heart with 3H-DOG and the recovery of intact mitochondria. The latter is determined using citrate synthase and this measurement might also be a useful refinement to the published NAD+ technique. As illustrated in Fig. 4, using the 3H-DOG entrapment technique, we have been able to confirm that the MPTP remains closed during 30 min of ischaemia but opens upon reoxygenation with a time course that reflects the return of pHi from its ischaemic value of <6.5 to preischaemic values [19–21].

Fig. 4

Time dependence of MPTP opening during reperfusion. The opening of the MPTP was detected using the DOG preloading technique. Hearts were subjected to 30 min of global ischaemia before reperfusion for the time shown. Data are taken from Ref. [21]. Parallel data for the pH of the perfusate are taken from Ref. [22].

Fig. 3

The use of [3H]6-deoxyglucose to measure MPTP opening in the perfused heart.

A potential advantage of the 3H-DOG entrapment technique over the NAD+ technique is that opening of the MPTP during isolation will release any of the 3H-DOG-6P that accumulated within the mitochondrial matrix as a result of pore opening in situ. Thus, it should only detect mitochondria undergoing the permeability transition in situ rather than during isolation. However, the DOG technique is not without its own limitations. The most significant of these is its inability to detect mitochondrial pore opening in cells that have undergone extensive necrosis [22]. The permeability barrier posed by the plasma membrane of necrotic tissues is compromised and mitochondrial integrity is totally destroyed. Thus, all the 3H-DOG that might have been entrapped in the cytosol and mitochondria will be lost and not detected, leading to a major underestimate of pore opening. Since the NAD+ technique relies on measuring the loss of mitochondrial NAD+, it does not suffer this disadvantage. However, the 3H-DOG technique does detect cells that have not yet progressed to full necrosis but are on their way and already exhibit compromised mitochondrial function through opening of the MPTP [22]. The extent to which myocytes have already undergone necrosis can be detected by classical techniques such as enzyme release.

3.2.1 The MPTP can open transiently during reperfusion with implications for apoptotic damage

Using the 3H-DOG entrapment technique, we were able to demonstrate that some opening of the MPTP could be detected during reperfusion after relatively short periods of ischaemia (10–20 min) even though total recovery of LVDP and ATP/ADP ratio was observed [19,20,23]. One explanation for this paradox would be that opening of the MPTP was transient, and rapidly followed by resealing that would enable total recovery of mitochondrial function and heart performance. Unfortunately, if resealing of the MPTP does occur, the entrapped 3H-DOG-P will remain within the mitochondria and thus closure will not be detected. To overcome this problem, hearts are loaded with 3H-DOG after a period reperfusion to allow functional recovery. If closure of pores does occur during reperfusion, mitochondrial 3H-DOG-P entrapment following postloading of 3H-DOG should be less than with preloading. Using this postloading technique, we demonstrated that 3H-DOG-P entrapment was only about 50% of that seen with preloading, confirming that some pore closure had occurred. Indeed, the extent of subsequent MPTP closure correlated with the functional recovery of the heart [21]. Thus, it would seem that if the insult caused by ischaemia/reperfusion is not too great, mitochondria may undergo a transient permeability transition, followed by pore closure and entrapment of the 3H-DOG-P. The closure probably occurs as a result of the loss of matrix [Ca2+] through the open pores and its subsequent removal from the cytosol. However, this will only occur if enough “healthy” mitochondria (without open pores) remain in the cell to accumulate the Ca2+ released by those with open pores. These healthy mitochondria will also provide sufficient ATP to maintain the ionic homeostasis of the cell. The balance between the number of “closed” and “open” mitochondria within any cell will be critical in determining whether a cell lives or dies. If there are too many “open” mitochondria, they will release more calcium and hydrolyse more ATP than the “closed” mitochondria can accommodate. In contrast, if there are sufficient “closed” mitochondria to meet the ATP requirements of the cell and also to accumulate released calcium without undergoing the permeability transition themselves, the “open” mitochondria will close again and the cell will recover, at least in the short term.

However, transient pore opening may have longer-term implications for the heart through inhibition of apoptosis that are not apparent over the time scale of perfusion experiments through the initiation of apoptosis. MPTP opening is associated with mitochondrial swelling, outer membrane rupture, and the release of proapoptotic factors such as cytochrome c from the intermembrane space. Once released, cytochrome c activates caspase 9, which in turn activates caspase 3. This protease mediates the proteolytic cleavage of a range of proteins responsible for the rearrangement of the cytoskeleton, plasma membrane, and nucleus that are characteristic of apoptosis [24,25]. A major distinction between apoptosis and necrosis is that the former requires ATP whilst the latter occurs in its absence [24,25]. Only if the MPTP opens sufficiently to cause cytochrome c release but then closes again to ensure that cellular ATP concentrations are maintained will the cell undergo apoptosis. Too much MPTP opening and necrosis will occur; too little opening and the cell recovers completely. In this sense, the mitochondrion is acting as the judge and executioner, the extent and duration of MPTP opening determining whether or not the cell lives or dies and the means by which it dies. It may be significant that the area around the necrotic core of an infarct shows a ring of apoptotic cell death [26,27]. Thus, in the core of the infarct, all cells are necrotic, reflecting permanent MPTP opening and mitochondrial disruption, whereas at the periphery, where ischaemic insult is less severe, transient MPTP opening may lead to apoptosis. This is illustrated in Fig. 5.

Fig. 5

Scheme illustrating how the extent of MPTP opening and subsequent closure may determine whether cardiac myocytes die by necrosis, as in the centre of an infarct, or apoptosis, as at the periphery. Further details may be found in the text.

4 The MPTP as a target for protecting hearts from reperfusion injury

It would be predicted that if opening of the MPTP is a critical factor in the transition from reversible to irreversible reperfusion injury of the heart, inhibitors of pore opening should offer protection. There is increasing evidence that almost any procedure that reduces reperfusion injury is associated with either a decrease in MPTP opening, or an increase in subsequent pore closure. This effect may be mediated either through direct inhibition of the pore with agents such as CsA and SfA, or through an indirect effect associated with a decrease in the factors responsible for MPTP opening such as oxidative stress and mitochondrial calcium overload. Recent data using H2O2-treated isolated cardiac myocytes have suggested that a priming phase, associated with mitochondrial swelling and cristae loss, but not depolarisation, may precede MPTP opening and be a target for myocardial protection [17]. The putative mitochondrial KATP channel opener, diazoxide (see Section 4.6), as well as the ANT inhibitor BKA (see Section 4.2) have been reported to inhibit this priming phase [28], although the mechanisms involved remain obscure and are not easily reconciled with recent data from this laboratory on the mitochondrial matrix volume in the ischaemic heart and reperfused heart [29].

4.1 Targeting CyP-D with Cyclosporin A and Sanglifehrin A to inhibit the MPTP

Nazareth et al. [30] were the first to demonstrate protection with CsA using an isolated cardiac myocyte model of anoxia and reoxygenation, and, subsequently, this has been confirmed by others [31,32]. Interestingly, in such cells, it has been shown that there is a correlation between mitochondrial [Ca2+] content and subsequent cell death [33,34]. In this laboratory, we were able to demonstrate protection by CsA in the Langendorff-perfused heart model of reperfusion injury [20,23] and, subsequently, by SfA [6,22]. Thus, as illustrated in Fig. 6, in the presence of 0.2 μM CsA or SfA, the recovery of haemodynamic function during reperfusion was greatly improved, as reflected by higher left ventricular developed pressure (LVDP) and lower end diastolic pressure (EDP) (an indicator of contracture whose elevation reflects elevated [Ca2+]), whilst greatly reduced release of intracellular lactate dehydrogenase confirmed that there was less necrotic cell damage. In addition, the CsA-treated hearts exhibited higher ATP/ADP ratios and lower AMP levels [6]. More recently, it has been demonstrated that CsA can significantly reduce infarct size in a coronary occlusion model of reperfusion injury, even when added only at reperfusion [35]. Taken together, these data strongly support the direct inhibition of the MPTP by CsA and SfA as being an effective means of inhibiting reperfusion injury. Thus, it is somewhat surprising that in experiments in which these agents produced significant functional protection of hearts from reperfusion injury, the 3H-DOG technique showed only a modest reduction in MPTP opening [22]. In contrast, under similar conditions, the use of mitochondrial NAD+ content as an indicator of MPTP opening showed a greater inhibitory effect of CsA [18]. However, DiLisa et al. did not account for the presence of broken mitochondria in their mitochondrial fraction. Since we have shown that CsA and SfA significantly increase the recovery of citrate synthase in the mitochondrial pellet, this may account for at least some of the differences between the two techniques. Indeed, when the better recovery of intact mitochondria from hearts treated with CsA and SfA was accounted for using the recovery of citrate synthase activity in the mitochondrial fraction, these agents were found to give a 35–50% decrease in 3H-DOG entrapment, more in line with the data obtained with the NAD+ technique [22]. Another important feature of the 3H-DOG entrapment technique is that it measures only MPTP opening that occurs in situ, whereas, as outlined above (Section 3.2), the mitochondrial NAD+ content may also reflect breakage of mitochondria as a result of MPTP opening during their isolation, and this is thought to be less in the CsA-treated hearts [6,19,22].

Fig. 6

Protection of hearts from reperfusion injury by Cyclosporin A and Sanglifehrin A. Note that 3H-DOG entrapment was measured using the preloading protocol. Data are taken from Refs. [6,21] where further details may be found.

We have recently recognised another factor that contributes towards the lack of correlation between the protective effects of CsA and SfA on the recovery of haemodynamic function and MPTP opening measured with the 3H-DOG entrapment technique [22]. A major limitation in the use of CsA and SfA as inhibitors of the MPTP is that both agents fail to inhibit pore opening when mitochondria are exposed to a sufficiently strong stimulus [6,19]. The elevated matrix calcium, oxidative stress, and adenine nucleotide depletion that accompany reperfusion after a period of ischaemia provide just such conditions [4,36]. However, as reperfusion continues and intracellular [Ca2+] and reactive oxygen species decline again, SfA and CsA may be able to block the MPTP totally, leading to better recovery of mitochondrial function than in control hearts. This is reflected in improved haemodynamic function (LVDP and EDP) and less necrotic damage (LDH release and reduced infarct size). Without such protection, more cells will progress towards necrosis, leading to the disruption of the plasma membrane, loss of ionic homeostasis, and release of LDH and cytosolic 3H-DOG-P. The resulting massive influx of calcium into these cells will then cause all their mitochondria to undergo the permeability transition and release their entrapped 3H-DOG-P. Hence, totally necrotic cells will lose both cytosolic and mitochondrial 3H-DOG-P, and so the presence of open mitochondria in such cells will not be not detected [22]. This may explain why the postloading technique fails to detect a decrease in pore opening with SfA and CsA [22] despite the improvement in heart function and decrease in LDH release [6,22]. It must also be recognised that factors other than disruption of mitochondrial function may lead to impaired haemodynamic function of the heart (stunning) in otherwise healthy myocytes. Thus, regimes that decrease stunning will cause an improvement in haemodynamic function without a measurable decrease in pore opening.

4.1.1 Limitations in the use of CsA

It is important to note that in addition to the limitations discussed above, there are two other major drawbacks to the use of CsA. First, it can potentially exert additional undesirable effects on the heart through inhibition of calcineurin-mediated processes [37]. This drawback can be overcome by the use of CsA analogues that are without effect on calcineurin such as [MeAla6]CsA and 4-methyl-val-CsA and by SfA [6,18,19]. The second problem is that CsA is only protective within a narrow concentration range, the optimal concentration of CsA for protection being about 0.2 μM with protection, declining again at higher concentrations [18,19,23]. This decline may partly reflect inhibition of calcineurin-dependent processes in the heart, but may also involve the emerging role of cyclophilins in the response of cells to oxidative stress. Thus, Doyle et al. [38] used antisense technology to knock out cytosolic cyclophilin (CyP-A) in cardiac myocytes and demonstrated that these cells were more sensitive to oxidative stress than control myocytes, but exhibited greater protection by CsA. Furthermore, overexpression of CyP-D has been shown to increases the resistance of cells to oxidative stress-induced cell damage [39]. There is a paradox here in that CyP-D is a component of the MPTP, and thus its overexpression might be expected to exacerbate damage, rather than offer protection. However, CyP-A has been shown to bind to peroxiredoxin IV and activate its peroxidase activity that uses thioredoxin to reduce H2O2 [40]. In addition, thioredoxin binds specifically to a cyclophilin in chloroplasts similar to CyP-D and reduces an intramolecular disulphide bond formed on oxidative stress [41]. Thus, it seems likely that CyP-D may play a dual role, protecting the mitochondria from oxidative stress initially, but if the insult becomes overwhelming, protection is reversed as the MPTP is activated and cell death ensues.

4.2 Targeting the MPTP with bongkrekic acid and ubiquinone derivatives

It is well established that BKA is a potent inhibitor of the MPTP in isolated mitochondria (see Section 2.2) and there are many reports of its use to inhibit pore opening in cultured cells (see Refs. [3,5]) including cardiac myocytes [15,28,42]. However, the primary action of BKA is to inhibit the export of ATP from the mitochondria to the cytosol and thus it is not appropriate for use in the heart, which is dependent on oxidative phosphorylation and the export of ATP from the mitochondria into the cytosol to drive contraction. The same considerations apply when using atractyloside to demonstrate a role for MPTP opening in reperfusion injury and studies where this has been attempted should be treated with caution [35,43,44]. Some ubiquinone derivatives have been described as potent inhibitors of the MPTP [9], and we have investigated the action of Coenzyme Q0 (2,3-dimethoxy-5-methyl-p-benzoquinone) on the Langendorff-perfused heart (Samantha Clarke, unpublished data). Unfortunately, we found that at 50 μM, the concentration reported to give maximal inhibition of the MPTP [9], this compound caused the heart to stop beating almost immediately. Even at 10 μM, a concentration that has relatively little effect on the MPTP, the performance of the heart was impaired, especially following ischaemia and reperfusion.

4.3 Na+/H+ antiporter inhibitors may target the MPTP by reducing calcium load and maintaining low intracellular pH during reperfusion

A potent inhibitor of MPTP opening is low pH (see Section 2.2), and several groups have demonstrated that maintaining an acidic extracellular pH during reoxygenation after a period of anoxia can protect cells from damage (see Ref. [45]). This is consistent with the observation that the MPTP opens during reperfusion only after the low pHi of ischaemia is restored to normal [21]. In contrast, acid pH during the ischaemic phase is detrimental, probably because Na+/H+ exchange is stimulated, loading the heart with Na+ and Ca2+. Inhibitors of the Na+/H+ exchanger such as amiloride and cariporide are thought to protect hearts from reperfusion injury by reducing the accumulation of Na+ and Ca2+ during the ischaemic phase (see Refs. [46,47]). However, there may be an additional effect during reperfusion associated with a slower return of pHi to normal physiological values (see Refs. [48,49]). Either mechanism would ultimately inhibit the opening of the MPTP. However, in preliminary experiments using the 3H-DOG entrapment technique, we have been unable to demonstrate any inhibition of MPTP opening during reperfusion by either amiloride or cariporide added prior to and during ischaemia and reperfusion (Samantha Clarke, unpublished data). Nevertheless, both agents did give protection under these conditions whether measured by haemodynamic function or lactate dehydrogenase release. As noted for CsA and SfA (Section 4.1), this does not rule out the inhibition of MPTP opening, but rather may indicate that the major effect is seen in the number of cells undergoing necrosis that are not detected by the 3H-DOG technique.

4.4 Targeting the MPTP with free radical scavengers including pyruvate and propofol

Reducing oxidative stress through the use of free radical scavengers is known to offer some protection against reperfusion injury [50]. Such agents may directly inhibit the opening of the MPTP by preventing oxidative cross-linking of critical cysteines on the ANT (see Ref. [5]). However, there may also be indirect effects on MPTP opening since oxidative stress is known to inhibit plasma membrane ion pumps, leading to perturbation of ionic homeostasis and calcium overload [51]. We have investigated the effects of two free radical scavengers on MPTP opening, pyruvate [21] and the anaesthetic propofol [52], and, in both cases, we have shown that their cardioprotective effects are associated with diminished opening of the MPTP.

4.4.1 Pyruvate

The ability of pyruvate to protect hearts and other tissues against ischaemia/reperfusion anoxia/reoxygenation injury has been known for many years and may, in part, be mediated by its free radical scavenging activity [21,53,54]. In addition, pyruvate is a superior fuel for the heart during reperfusion because, unlike glucose or glycogen, it does not require ATP for activation before it can be metabolised. We have confirmed that 10 mM pyruvate added before ischaemia and maintained during reperfusion greatly improves the functional recovery of hearts, and that this is associated with a major reduction in MPTP opening during the initial stages of reperfusion. Furthermore, postloading experiments demonstrated that as reperfusion continued, almost all the “open” mitochondria subsequently resealed [21]. These data provide further evidence that once opened, the MPTP can close again and allow hearts to recover fully during reperfusion, provided the initial insult is not too great. During the course of these experiments, it became apparent that pyruvate also caused a greater accumulation of intracellular lactic acid during ischaemia and slowed the return of pHi during reperfusion to normal values from the low values of ischaemia [21]. There is also direct evidence from nuclear magnetic resonance (NMR) studies that pyruvate causes a decrease in pHi in a low-flow model of ischaemia [55]. In view of the inhibitory effect of low pH on the MPTP, this might provide yet another mechanism by which pyruvate can protect the heart against reperfusion injury.

4.4.2 Propofol

The anaesthetic propofol is frequently used during cardiac surgery and in postoperative sedation [56]. It acts as a free radical scavenger and may also inhibit plasma membrane calcium channels (see Ref. [52]). In addition, at concentrations higher than used clinically, propofol may inhibit the MPTP directly [57]. Propofol has been shown by others to protect the Langendorff-perfused heart against reperfusion injury [58] and damage caused by H2O2-induced oxidative stress [59]. We have used the 3H-DOG technique to confirm that this protection is accompanied by less opening of the MPTP in situ [52]. Furthermore, mitochondria isolated from the propofol-treated hearts exhibited less pore opening than control mitochondria exposed to the same [Ca2+]. However, when propofol was added to isolated heart mitochondria at the same concentration as used in the heart perfusion (2 μg/ml, a concentration typically used in clinical anaesthesia), no inhibition of MPTP opening was observed [52], suggesting that that the protective effect of propofol may be secondary to the decreased oxidative stress and calcium overload. We have also demonstrated a cardioprotective effect of propofol on the functional recovery of the working rat heart following cold cardioplegic ischaemic arrest, a model that is closer to the situation experienced in open heart surgery [52]. Most recently, we have extended these studies to an in vivo pig model of cardiopulmonary bypass with warm blood cardioplegia that closely matches current clinical practice. Here, too, normal clinical concentrations of propofol improve functional recovery of the heart, reduce troponin I release, and maintain higher tissue ATP levels [60].

Whatever the exact mechanisms involved, both pyruvate and propofol provide examples of reagents whose protection of the heart from reperfusion injury is accompanied by a decrease in MPTP opening in vivo. These data suggest that propofol and pyruvate may be a useful adjunct to the cardioplegic solutions used in cardiac surgery.

4.5 Elevated extracellular [Mg2+] may protect hearts by inhibiting the MPTP

Magnesium is well known to protect hearts from ischaemia and reperfusion injury and seems most effective when present at high concentrations (>8 mM) during the reperfusion phase [61–63]. The available evidence suggests that the protective effect involves extracellular Mg2+, since intracellular [Mg2+] increases very little even at greatly increased [Mg2+] in the perfusion medium [62,64]. It is generally accepted that it exerts its protective effects by inhibiting L-type calcium channels and the Na+/Ca2+ antiporter, thus decreasing calcium overload [61,62]. Indeed, other inhibitors of plasma membrane or mitochondrial calcium channels such as verapamil and ruthenium red derivatives have been shown to protect hearts from reperfusion injury in this manner (see Ref. [4]). There are also data to suggest that the presence of supraphysiological [Mg2+] prior to ischaemia exerts an antioxidant effect during reperfusion [63]. Taken together, these data would suggest that protection afforded by elevated [Mg2+] may also involve the inhibition of MPTP opening.

4.6 Protection by ischaemic preconditioning involves inhibition of MPTP opening

One of the most effective ways to protect the heart against reperfusion injury is to subject them to brief ischaemic periods with intervening recovery periods before the prolonged period of ischaemia is initiated. Such “ischaemic preconditioning” is associated with two phases of protection: an immediate effect and a “second window” that occurs 24–48 h later [65]. The exact mechanisms involved in preconditioning are still being debated, but several processes have been implicated. The longer-term effects are probably caused by stimulation of the transcription of specific genes, perhaps through a mechanism activated by free radicals and stress-activated protein kinases. Of particular interest may be the up-regulation of heat shock proteins since recent data have shown that heat shock specifically up-regulates the expression of liver mitochondrial Hsp25 and this is associated with desensitisation of MPTP opening to Ca2+ and HgCl2 in isolated mitochondria [66]. It is also known that heart mitochondria from mice in which Hsp25 is down-regulated are more sensitive to MPTP opening and also exhibit hallmarks of oxidative stress including oxidatively damaged ANT [67]. The mechanisms responsible for the short-term effects of preconditioning include the activation of protein kinase C (PKC). This may be mediated either by reactive oxygen species released during the short intervening reperfusion periods, or by factors released during the brief ischaemic periods such as adenosine, bradykinin, noradrenaline, and opioids. Thus, PKC inhibitors and free radical scavengers antagonise IPC, whilst adenosine agonists and PKC activators mimic the effect (see Refs. [65,68,69]). The ultimate target of these kinases is unknown, although it may be significant that activation of PKCε and its translocation to mitochondria has been reported to be important for preconditioning [70,71]. There is also evidence for an involvement of sulphonylurea-sensitive KATP channels, since KATP channel openers such as diazoxide can mimic IPC whilst blockers such as glibencamide inhibit [72,73]. Furthermore, PKC-dependent activation of plasma membrane KATP channels by IPC has been demonstrated [74].

As we have noted previously [2], if the extent of MPTP opening is a critical factor in determining the extent of reperfusion injury, it would be predicted that IPC should reduce the amount of pore opening. Indirect evidence obtained using isolated cardiac myocytes and mitochondria has implicated the inhibition of the MPTP in both calcium- and diazoxide-mediated preconditioning [32,35,75]. We have recently used the 3H-DOG technique to demonstrate directly that IPC inhibits the MPTP in the Langendorff-perfused heart [22]. IPC not only reduced the opening of the MPTP during the early phase of reperfusion (measured with 3H-DOG preloading) but also increased subsequent pore closure (measured with 3H-DOG postloading) as illustrated in Fig. 7. However, MPTP opening in isolated mitochondria from preconditioned hearts showed a greater sensitivity towards calcium than mitochondria from control hearts [22], implying that the inhibition of MPTP opening in situ by IPC is probably indirect. It may well be mediated either through a decrease in calcium loading of the cardiac myocyte or a decreased production of ROS upon reperfusion, both of which are observed in response to ischaemic and diazoxide-mediated preconditioning (see Ref. [22] for detailed references). Indeed, our own data suggest that such an indirect mechanism of inhibiting the MPTP may be more effective at protecting hearts from reperfusion injury rather than targeting the MPTP directly with CsA or SfA [22]. This may reflect the inability of CsA and SfA to inhibit the MPTP when matrix calcium concentrations and oxidative stress are high as may occur in the initial phase of reperfusion [19,76].

Fig. 7

Protection of hearts from reperfusion injury by ischaemic preconditioning is accompanied by decreased MPTP opening. Two 5-min ischaemic periods with intervening recovery were used for preconditioning before 30 min of global isothermic ischaemia and then reperfusion. Mitochondrial 3H-DOG entrapment was measured in nonischaemic hearts and following reperfusion using both the preloading and postloading protocols as indicated. Error bars represent the S.E. for eight observations. Data are taken from Ref. [22] where further details may be found.

In contrast to our own data suggesting that IPC exerts its effects on the MPTP indirectly, Baines et al. [44] have very recently provided evidence that PKCε may interact directly with VDAC to cause its phosphorylation and consequent inhibition of MPTP opening. However, there are aspects of this study which need clarification, especially the quantification of the inhibition of MPTP opening in isolated mitochondria and the ability of added PKCε to cause inhibition in the absence of added ATP. Furthermore, the in vitro phosphorylation of VDAC was performed using a glutathione-S-transferase fusion protein in free solution and without either detergent or added phospholipids. Under these conditions, the conformation state of VDAC, which is normally embedded in the outer membrane, may well be far removed from its native state.

4.6.1 Mitochondrial KATP channels and ischaemic preconditioning

It was originally proposed that the opening of the plasma membrane KATP channel might hyperpolarize the cell and lead to a shorter action potential duration (APD) and calcium loading. However, data obtained with a range of KATP channel openers showed a poor correlation between their effect on APD and their protective effects and, in recent years, the emphasis has shifted towards a role for the putative mitochondrial KATP channel in IPC [72,77]. One reason for this is that diazoxide protects hearts at concentrations much lower than reported to open the plasma membrane KATP channel, but similar to values reported for the mitochondrial KATP channel. Furthermore, protection was inhibited by 5-hydroxydecanoate (5-HD), an agent reported to be a specific inhibitor of the mitochondrial KATP channel (but see below). However, on several counts, these data are not totally convincing. First, there are reports that diazoxide can work at much lower concentrations on the plasma membrane KATP channel when ADP is present (as it will be in the cell) [78]. Second, the opening of a mitochondrial K+ channel would depolarize mitochondria (inhibiting oxidative phosphorylation) and induce K+ cycling (effectively increasing ATP demand). Yet published data suggest the opposite to be the case [79] as might be predicted for a protective agent. Third, recent reports have shown that in knockout mice lacking the plasma membrane potassium channel Kir6.2, neither IPC nor diazoxide was able to offer protection from reperfusion injury [80,81]. Fourth, diazoxide has been shown by several workers to inhibit succinate dehydrogenase activity and hence respiration supported by the citric acid cycle [29,82–84]. This inhibition of respiration might be responsible for the increased mitochondrial production of ROS observed with diazoxide treatment [85–87], with the ROS then mediating IPC [68,88]. Fifth, 5-HD cannot be used as a specific mitochondrial KATP channel inhibitor since it is a racemic mix of d-isoforms and l-isoforms of a substituted fatty acid that can be activated to its coenzyme A derivative and then act as either a substrate or inhibitor of fatty acid β-oxidation [29,83,89,90]. Sixth, in isolated heart mitochondria incubated under conditions that mimic the in vivo situation, we have been unable to demonstrate any effect of diazoxide, 5-HD, glibencamide, and a range of other putative mitochondrial KATP channel openers or blockers on light scattering or isotopic measurements of matrix volume [84].

These considerations have led us to question whether such mitochondrial KATP channels have any significance in regulating heart mitochondrial function [84]. That is not to say that mitochondrial potassium channels may not play an important role in regulating mitochondrial function under some conditions. Indeed, we have provided extensive evidence that there is a calcium-activated potassium channel in liver mitochondria that is involved in mediating the increase in matrix volume occurring in response to hormones such as glucagon and adrenaline. This plays an important role in the stimulation of the respiratory chain by these hormones [91]. Furthermore, we have shown that IPC and ischaemia itself can increase the matrix volume of heart mitochondria in situ, leading to a stimulation of respiratory chain activity that may represent an important adaptation to maintain rates of respiration during hypoxia [29]. However, this effect is not prevented by 5-HD, again implying that the proposed mitochondrial KATP channel is not involved. Others have very recently provided evidence for such a calcium-activate K+ channel in heart mitochondria that is distinct from the KATP channel [92].

5 Conclusions

The opening of the MPTP converts the mitochondrion from an organelle that provides ATP to sustain heart function into an instrument of cell death by apoptosis if the insult is mild, to necrosis if the insult is profound. Whatever the final pathway of cell death, pharmacological interventions that can inhibit MPTP opening and enhance pore closure, either directly (CsA and SfA) or indirectly (propofol, pyruvate, and preconditioning protocols), provide protection from reperfusion injury. In cardiac surgery, pharmacological agents can be administered prior to ischaemia, enabling them to exert their effects on mitochondria prior to ischaemia and reperfusion. However, in restoring flow to a blocked coronary artery, either with clot-busting enzymes or angioplasty, it will be more difficult to ensure that the drug is in place in the critical early stage of reperfusion. In a rat coronary ligation model, there is evidence that CsA can act when added at reperfusion that is promising [35], but the nonspecific effects of CsA and its narrow therapeutic window (see Section 4.1.1) represent a severe limitation. A major goal is the development of novel, specific, and potent inhibitors of the MPTP that can enter the heart rapidly and hopefully reach the ischaemic area from the collateral circulation.


This work was supported by project grants from the British Heart Foundation to A.P.H. and an International Exchange Fellowship from the Royal Society to S.A.J.


  • Time for primary review 14 days

  • Abbreviations:
    adenine nucleotide translocase
    bongkrekic acid
    Cyclosporin A
    end diastolic pressure
    ischaemic preconditioning
    left ventricular developed pressure
    mitochondrial permeability transition
    mitochondrial permeability transition pore
    peptidyl-prolyl cistrans isomerase
    reactive oxygen species
    Sanglifehrin A
    voltage-activated anion channel


  1. [1]
  2. [2]
  3. [3]
  4. [4]
  5. [5]
  6. [6]
  7. [7]
  8. [8]
  9. [9]
  10. [10]
  11. [11]
  12. [12]
  13. [13]
  14. [14]
  15. [15]
  16. [16]
  17. [17]
  18. [18]
  19. [19]
  20. [20]
  21. [21]
  22. [22]
  23. [23]
  24. [24]
  25. [25]
  26. [26]
  27. [27]
  28. [28]
  29. [29]
  30. [30]
  31. [31]
  32. [32]
  33. [33]
  34. [34]
  35. [35]
  36. [36]
  37. [37]
  38. [38]
  39. [39]
  40. [40]
  41. [41]
  42. [42]
  43. [43]
  44. [44]
  45. [45]
  46. [46]
  47. [47]
  48. [48]
  49. [49]
  50. [50]
  51. [51]
  52. [52]
  53. [53]
  54. [54]
  55. [55]
  56. [56]
  57. [57]
  58. [58]
  59. [59]
  60. [60]
  61. [61]
  62. [62]
  63. [63]
  64. [64]
  65. [65]
  66. [66]
  67. [67]
  68. [68]
  69. [69]
  70. [70]
  71. [71]
  72. [72]
  73. [73]
  74. [74]
  75. [75]
  76. [76]
  77. [77]
  78. [78]
  79. [79]
  80. [80]
  81. [81]
  82. [82]
  83. [83]
  84. [84]
  85. [85]
  86. [86]
  87. [87]
  88. [88]
  89. [89]
  90. [90]
  91. [91]
  92. [92]
View Abstract