The mitochondrial permeability transition pore: Molecular nature and role as a target in cardioprotection

Abstract

The mitochondrial permeability transition (PT) – an abrupt increase permeability of the inner membrane to solutes – is a causative event in ischemia–reperfusion injury of the heart, and the focus of intense research in cardioprotection. The PT is due to opening of the PT pore (PTP), a high conductance channel that is critically regulated by a variety of pathophysiological effectors. Very recent work indicates that the PTP forms from the F-ATP synthase, which would switch from an energy-conserving to an energy-dissipating device. This review provides an update on the current debate on how this transition is achieved, and on the PTP as a target for therapeutic intervention. This article is part of a Special Issue entitled "Mitochondria: from basic mitochondrial biology to cardiovascular disease".

Highlights

• •The mitochondrial permeability transition pore plays a key role in heart disease.
• •Existing models for the permeability transition pore are critically reviewed.
• •F-ATP synthase is the best molecular candidate for pore formation.
• •Cyclophilin D is an important pore regulator but not a structural component.
• •Cyclophilin inhibitors are promising but promiscuous drugs.

1. Introduction

The permeability transition (PT) is an abrupt increase of the inner mitochondrial membrane (IMM) permeability to solutes, which in mammalian mitochondria has a cutoff of about 1500 Da. Occurrence of the PT and its inhibition by adenine nucleotides is known since the 1950s [1,2], and the phenomenon has been investigated in a number of laboratories (e.g. [3–13]). The term “permeability transition” was introduced in 1979 by Haworth and Hunter, who carried out a thorough characterization of its basic features in heart mitochondria, and provided the important insight – which is today generally accepted – that the PT could be due to opening of an IMM channel, the PTP [14–17]. This hypothesis was confirmed by patch-clamp studies on mammalian mitoplasts, which revealed the presence of a high-conductance (≈ 1 nS) channel, the mitochondrial megachannel (MMC) [18,19]. The MMC possesses all the basic features of the PTP [20,21] including sensitivity to cyclosporin A (CsA) [22], and represents the electrophysiological equivalent of the pore [23]. The study of mitochondrial channels has greatly contributed to our understanding of mitochondrial physiology, and to the acceptance of the pore theory of the PT (see [24] for a recent review).

PTP opening is traditionally linked to mitochondrial dysfunction because its occurrence leads to mitochondrial depolarization, cessation of ATP synthesis, Ca^2 + release, pyridine nucleotide depletion, inhibition of respiration and, in vitro at least, matrix swelling; in turn, swelling causes mobilization of cytochrome c, outer mitochondrial membrane (OMM) rupture and eventually release of proapoptotic proteins such as cytochrome c itself, endonuclease G and AIF [25,26]. It should be mentioned that these detrimental effects on energy conservation and cell viability are only seen for long-lasting openings of the PTP [27], while short-term openings – which have been documented both in isolated mitochondria and in situ [27–30] – may be involved in physiological regulation of Ca^2 + and reactive oxygen species (ROS) homeostasis [31], and provide mitochondria with a fast mechanism for Ca^2 + release [32–35]. The potential role of the PTP in heart injury has been recognized very early [36,37], well before the role of mitochondria in apoptosis was discovered [38–40]. PTP desensitization with CsA proved beneficial in heart ischemia–reperfusion injury, as well as in pre- and post-conditioning through mechanisms that await clarification [41–49].

Matrix Ca^2 + is an essential permissive factor for PTP opening, but the role of mitochondrial “Ca^2 + overload” as a causative event in I/R injury of the heart has recently been challenged. In MCU null mitochondria – where Ca^2 + overload does not occur during reperfusion – the extent of necrosis was the same as that observed in the hearts from wild type littermates, and the cardioprotective effect of CyPD ablation was abrogated [50]. These surprising observations raise many issues that still await an answer, such as the cause of cell death, the mechanism of activation of mitochondrial metabolism and the mechanism of PTP opening in MCU null mice. Yet these experiments do show that cardiomyocyte cell death can occur without mitochondrial “Ca^2 + overload”; and suggest that there is enough Ca^2 + in the matrix of MCU null mitochondria to allow pore PTP opening, possibly a consequence of the burst of ROS that follows reperfusion [31].

2. Molecular nature of the permeability transition pore: the early days

The molecular nature of the PTP has been the matter of debate for the last 30 years. In the early 1990s Snyder and coworkers found that the peripheral benzodiazepine receptor, an OMM protein today called TSPO [51], copurified with the adenine nucleotide translocator (ANT) and the voltage-dependent anion channel (VDAC) in protocols based on detergent extraction followed by hydroxylapatite chromatography; radiolabeled high-affinity ligands of TSPO were recovered in fractions where TSPO could be detected together with VDAC and ANT [52]. This finding was of great interest because nanomolar concentrations of the same TSPO ligands affected the channel properties of MMC in electrophysiological experiments, suggesting that all of these proteins could be involved in formation of the PTP [53].

This suggestion was strengthened a few years later by work from the Brdiczka laboratory during the characterization of OMM and IMM “contact sites”, i.e. specialized structures where the two membranes form close contacts mediated by protein–protein interactions [54]. These sites would include hexokinase on the cytosolic surface of, and VDAC within, the OMM, creatine kinase and nucleoside diphosphate kinase in the intermembrane space, and ANT in the IMM; they were proposed to mediate channeling of adenine nucleotides to and from mitochondria [54–56]. The link with the PTP was made when the same laboratory showed that hexokinase-enriched fractions from low detergent extracts of mitochondria formed channels with the conductance expected of the PTP, and conferred permeability properties to liposomes that could be inhibited by N-methylVal-4-cyclosporin [57]. It must be stressed that the preparation contained a very large number of proteins, which makes assignment of the channel activity to a specific species quite problematic. Furthermore – and unlike the case of PTP – currents were inhibited rather than induced by atractylate, and the active fractions were not enriched in VDAC and/or ANT [57]. The same preparations were shown to also contain proteins of the Bcl-2 family [58], and this set of observations led to a model where the PTP would be a multiprotein complex spanning both mitochondrial membranes and comprising ANT, VDAC, TSPO, cyclophilin (CyP) D as well as hexokinase and Bcl-2 proteins [59]. This model did not stand the test of genetics, as a CsA-sensitive PT could be easily detected in the absence of ANT [60], VDAC [61,62] as well as of TSPO [63]. An alternative model is the formation of the PTP by the Pi carrier following its interaction with CyP-D and ANT [64]. However, results obtained by patch-clamp analysis of the reconstituted Pi carrier do not match the electrophysiological PTP features [65] and genetic deletion of the Pi carrier does not support the idea that this protein is essential for PTP formation [66].

Studies on Ppif^−/− mice (Ppif is the unique gene encoding CyPD in the mouse) have demonstrated that this protein is an important modulator which sensitizes the PTP to Ca^2 + and confers sensitivity to CsA, but not an essential pore component [67–70]. By following the interactions of the matrix CyPD with other mitochondrial proteins it has recently been possible to identify a novel structure for the PTP, which will be described in the following paragraph.

3. The permeability transition pore forms from F-ATP synthase

By monitoring the presence of CyPD in blue native gels of mitochondrial proteins Giorgio et al. discovered that CyPD interacts with the F-ATP synthase, and that it can be crosslinked to the stalk proteins b, d and OSCP [71]. Binding of CyPD to the F-ATP synthase required Pi, and caused a decrease of the enzyme's catalytic activity; while it was counteracted by CsA, which displaced CyPD and increased the catalytic activity [71]. It was then found that CyPD interacts with the OSCP subunit of F-ATP synthase [72]. Gel-purified dimers of F-ATP synthase incorporated into lipid bilayers displayed currents activated by Ca^2 +, Bz-243 and phenylarsine oxide (but not atractylate) with a unit conductance of about 500 pS, which is identical to that of the bona fide mammalian MMC-PTP [72]. The channel-forming property is shared by purified F-ATP synthase dimers of yeast mitochondria, which also displayed Ca^2 +-dependent currents of slightly lower conductance (about 300 pS) [73]. Furthermore, yeast strains lacking the e and/or g subunits, which are necessary for dimer formation, showed a remarkable resistance to PTP opening [73]. Although strains lacking subunits e [74] or g [75] display abnormal morphology, with balloon-shaped cristae and F-ATP synthase monomers distributed randomly in the membrane, they did develop a normal membrane potential [73], suggesting that the increased resistance to PTP opening may not depend on these structural differences. Based on these findings, it has been proposed that the PTP forms from F-ATP synthase dimers, possibly in the lipid region between two adjacent stalks [76].

The idea that the pore forms from the F-ATP synthase is also supported by two independent studies. Bonora et al. used targeted inactivation of the c subunit of F-ATP synthase – which forms the H⁺-transporting c ring of F-ATP synthases – to show that HeLa cells become resistant to PTP opening and cell death [77]; while Alavian et al. reconstituted the c subunit or the purified F-ATP synthase in liposomes, and measured Ca^2 +-activated channels [78] with properties similar to those described by Giorgio et al. with purified dimers [72]. It is not possible to derive mechanistic insights about the nature of the PTP-forming channel from the study of Bonora et al. because the consequences of knockdown of the c subunit on other components of the F-ATP synthase and on other mitochondrial proteins were not addressed, and it is unclear whether and how many functional F-ATP synthases were left after the knockdown of the c subunit [77]. Alavian et al., on the other hand, suggested that the channel of the PTP forms within the c ring itself after Ca^2 +-dependent extrusion of F₁, i.e. of the γ subunit [78]. We think that this hypothesis is extremely unlikely for the following reasons:

• •Displacement of F₁ from F_O requires very drastic conditions, such as treatment with 2 M urea [79] yet a functional F_OF₁ complex can be easily reconstituted after treatment with urea, indicating that the γ/δ/ε subunit reinserts into F_O. It is hard to envision a plausible mechanism through which matrix Ca^2 + could cause release of F₁, and then create within F_O a channel that cannot be closed by subunit γ/δ/ε [78].
• •Alavian et al. reported that the “F_O channel” can instead be closed by the β subunit, and suggested that this is the mechanism through which pore closure occurs in situ [78]. There are major problems with this proposal, because structural studies have established that subunit β does not interact with the c ring [80]; and it is not obvious where the free β subunit would come from, given the extreme resistance of the F₁ subcomplex to denaturation. This hypothesis is also difficult to reconcile with the well established fact that PTP–MMC opening is readily and fully reversible upon chelation of Ca^2 + in mitoplasts [21], intact mitochondria [81] as well as in reconstituted dimers of F-ATP synthase [72].
• •Channel openings were also seen with preparations of the whole F-ATP synthase, and these could be inhibited by CsA after the addition of Ca^2 + [78]. If the mechanism of pore opening is “expulsion” of F₁ by Ca^2 +, it is not easy to explain how the current could be inhibited by CsA. Indeed, it is firmly established that CsA inhibits the pore by removing CyPD, which in turn interacts with F₁, not F_O [71,72].
• •F₁ has binding sites that can accommodate directly the effects of Ca^2 +, Mg^2 +, adenine nucleotides and Pi; and through CyPD (un)binding those of H⁺, CsA and possibly of oxidants [82]. Any model of the pore must account for all inducers and inhibitors, which appear difficult to fit in the c ring, a multimer of identical c subunits.
• •Silencing of the ATP5E gene, which encodes the ε subunit, resulted in downregulation of the F-ATP synthase complex with accumulation of subunit c, yet mitochondria were more coupled [83], which is the opposite of what would be expected if the isolated c ring can form membrane channels.
• •McGeoch and coworkers have performed patch-clamp studies of highly purified c subunit (which gave a single silver-stained band and was validated by sequencing) with very different results, as currents were inhibited rather than activated by Ca^2 + [84–86]. It is legitimate to wonder whether other F-ATP synthase components were present in the preparation of Alavian et al. that could explain this discrepancy.

Given that in our hands the PTP-MMC readily forms from F-ATP synthase dimers but not monomers [72]; and that inactivation of the “dimerization” subunits e and g in Saccharomyces cerevisiae increases resistance of the PTP to Ca^2 + [73], we favor the idea that the pore forms at the interface between two monomers in the dimeric enzyme, as we will discuss further after covering the regulatory role of CyPD.

4. Modulation by cyclophilin D and cyclosporin A

CyPs are ubiquitous, conserved proteins possessing peptidyl prolyl cis-trans isomerase activity [87–89]. Sixteen isoforms have been found in man, and the most abundant (and the first to be discovered) is cytosolic CyPA [90]. The enzymatic activity of all CyPs is inhibited by CsA [91] and the CsA/CyPA complex inhibits the cytosolic phosphatase calcineurin [92]; as a result, NFAT is no longer dephosphorylated, an event that prevents its nuclear translocation causing immunosuppression [93,94].

Mammals possess a unique mitochondrial species called CyPD, which in the mouse is encoded by the Ppif gene (see [95] for a review). CyPD is the mitochondrial receptor for CsA and modulates the PTP, but it is not a structural pore component. Indeed, the PTP can still open in mitochondria from Ppif^−/− mice, although higher Ca^2 + loads are required [67–70]. Regulation by CyPD may be a relatively recent evolutionary event [96], since the PTPs of S. cerevisiae and D. melanogaster are insensitive to CsA [97,98]. As also discussed elsewhere, the effect of CsA on the mammalian PTP is best described as “desensitization” in the sense that the PTP can still occur but becomes more resistant to Ca^2 +, Pi and other inducers [25,76]. This consideration is important because (i) CsA can desensitize but not block the PTP, and therefore lack of sensitivity to CsA does not necessarily imply that the PTP is not involved in the event being studied; and (ii) different cells express different levels of CyPD, and obviously only CyPD-expressing mitochondria can respond to CsA [99].

CyPD binds the lateral stalk of F-ATP synthase (OSCP, b and d subunits) [71]. Like PTP induction, binding requires Pi (and results in partial inhibition of ATP synthase activity); while CsA displaces CyPD resulting in enzyme reactivation [71]. We have recently identified the binding site of CyPD to the OSCP subunit, possibly in a region comprising helices 3 and 4 [72], which is also the binding site of Bz-423 [100,101]. In keeping with PTP formation by F-ATP synthase, decreased levels of OSCP halved the threshold Ca^2 + load required for PTP opening [72].

As discussed in detail elsewhere [72,76] we assume that the PTP forms within the IMM at the interface between two adjacent F_O sectors. Matrix Ca^2 + would have an essential permissive role in PTP formation after binding to the catalytic Me^2 +-binding site, which is usually occupied by Mg^2 +, and could be influenced by OSCP. Our working hypothesis is that OSCP as such is a “negative” modulator, whose effect can be counteracted by binding of the “positive” effector CyPD (which indeed decreases the threshold Ca^2 + required for PTP opening). Removal of OSCP, or CyPD binding to OSCP, would induce similar conformational effects on the rigid stalk proteins, leading to increased probability of PTP opening at the IMM — a working hypothesis that awaits experimental testing.

We note that PTP formation at the membrane interface between two stalks could also accommodate the PTP-modulating effects of fatty acids [5–7,102]. The role of cardiolipin should also be explored, as it stabilizes respiratory supercomplexes [103] and, due to its partitioning into high-curvature membrane regions, plays a role in cristae formation and morphology [104]. Indeed, cardiolipin increases the degree of oligomerization of F-ATP synthase by promoting the formation of extended dimer rows, which is compromised in D. melanogaster mutants defective for cardiolipin synthesis [104]. The high susceptibility of cardiolipin to oxidation might alter F-ATP synthase conformation, affecting in turn the PTP open probability.

5. Regulation of the permeability transition by the outer mitochondrial membrane

The PT is an inner membrane event, since it also occurs in mitoplasts, i.e. mitochondria stripped of the OMM [105], yet the OMM does play a role in pore modulation. Lê-Quôc and Lê-Quôc were the first to show that induction of the PTP by substituted maleimides requires the OMM [106]. We have confirmed that PTP opening by N-ethylmaleimide is no longer present in mitoplasts [63] and extended the “sensitizing” role of the OMM to protocols where the PTP is modulated by dicarboxylic porphyrins plus visible light, a treatment that leads to the efficient production of singlet oxygen [107]. Low light doses inactivate the PTP through degradation of hystidyl residues, which in turn prevents matrix cysteine oxidation [108], possibly through a conformational change; while higher light doses activate the PTP through OMM cysteine oxidation [109]. PTP activation strictly required an intact OMM, since the inducing effect of hematoporphyrin plus light was completely lost in mitoplasts [105]. Largely based on the effect of its ligands on the PTP, it was proposed that the OMM protein responsible for PTP sensitization was the peripheral benzodiazepine receptor (today called TSPO) [52,53,105,110–112], but this turned out to be incorrect.

It had been known for some time that many cellular effects of “TSPO ligands” were not due to an interaction with TSPO, suggesting the existence of a different site of action [113,114]. The relevant target for PTP modulation, however, was identified only recently by 2 independent sets of observations. The Glick laboratory discovered that “TSPO ligands” (including the benzodiazepine Bz-423 and PK11195) interact with the mitochondrial F-ATP synthase and affect its activity [100,101,115,116]; while Giorgio et al. demonstrated that Bz-423 mimics the PTP-inducing effects of CyPD, and is able to favor the Ca^2 +- and oxidant-induced transition of F-ATP synthase dimers to unselective high-conductance channels [72]. Thus, TSPO ligands directly modulate the propensity of F-ATP synthase to form channels; and it may therefore come as no surprise that mitochondria from mice with genetic ablation of TSPO (which completely lack high-affinity binding sites for PK11195) respond with PTP induction to PK11195, N-ethylmaleimide and photooxidative stress [63]. Thus, TSPO is not responsible for the PTP-modulating effects of “TSPO ligands”, and the protein(s) responsible for pore induction by N-ethylmaleimide and photooxidative stress await clarification. A potential candidate is Abcb6, an ATP binding cassette transporter of the OMM involved in heme and porphyrin homeostasis [117].

It is worth mentioning that the vast majority of F-ATP synthase dimers is located in rows of oligomers inside cristae [75,118,119] where a direct interaction with the OMM is not possible. Thus, either the PTP forms in the small population of dimers facing the intermembrane space, where direct contact with the OMM can occur; or the effect of the OMM is exerted by controlling the diffusion of PTP-regulating metabolites and ions, including Ca^2 + itself [120,121] in a process that would be greatly favored by, and contribute to, cristae remodeling [122].

The OMM may also affect the outcome of PTP opening without directly affecting its open-closed transitions. Indeed, it has been shown that the PT causes matrix swelling and OMM rupture only in the presence of Bax and Bak, whose genetic ablation confers OMM resistance to swelling, thus preventing organelle rupture and cell death [123].

6. Cyclosporin A and cardioprotection

Treatment with CsA confers remarkable protection against acute myocardial injury induced by post-ischemic reperfusion [37], and this is matched by genetic ablation of CyPD [67,69], which suggests that the protective effects of CsA depend on CyPD inhibition and, by inference, on PTP desensitization.

Cardioprotection by CsA has been described in a wide variety of experimental models and animal species including humans [124]. All in all, we think that CsA provided a terrific proof of concept about the causative role of the PTP in I/R injury of the heart, and as a target for cardioprotection. There may be exceptions, however, as lack of protection by CsA has been reported in in vivo I/R in rat [125] and pig hearts [126], while a recent report has instead confirmed the cardioprotective efficacy in pig hearts [127]. The basis for these discrepancies is not immediately obvious, but key elements could be the duration of ischemia, which determines severity of damage [128], and CsA dosage [37]. In an apparent paradox, in perfused hearts of Ppif^−/− (CyPD null) mice the extent of necrosis increased when reperfusion was established after a relatively short ischemic episode (30 min), while the expected protection was observed when ischemia was prolonged to 60 min [128]. Additional studies should address the factors involved in mild I/R damage that are not affected by treatments targeting CyPD.

Regarding dosage, it should be recalled that the protective effect of CsA is observed in a very narrow dosage window; this was first shown in perfused rat hearts where protection was seen at 0.2 but not 1 μM [37], a finding that has been confirmed by several laboratories, including ours, and never disputed. It seems possible that the effective CsA concentration may differ in different species, and therefore that conditions defined as optimal in one model may prove ineffective in others. The limitations discussed above, however, must be seriously taken into account when devising possible clinical applications for CsA, also considering the potentially adverse effects of prolonged PTP inhibition in failing hearts [129].

7. Additional effects of cyclophilin D and cyclosporin A

In the mouse CyPD is encoded by a unique nuclear gene (Ppif), and the transcript includes a matrix targeting sequence that is removed from the protein after mitochondrial import [130]. Consistently, the protein localizes to the mitochondrial matrix as detected both by electron microscopy and by trypsin titrations of mitochondrial fractions [131,132]. Protection afforded by deletion of CyPD or its inhibition is generally referred to its localization in the mitochondrial matrix, which allows regulatory interactions with F-ATP synthase and the PTP. CyPD has been proposed as a key interactor for the Hsp90–Trap-1 complex [133] and for p53 [134]. By sequestering a relevant fraction of CyPD these proteins would reduce the probability of PTP opening thus favoring survival of cancer cells, a provocative hypothesis that is the subject of controversy [135].

Genetic ablation of CyPD also caused mild ER stress, as judged on the basis of increased phosphorylation of eIF2a and expression of GRP78 without changes of the potentially detrimental CHOP [136]. This mild ER stress appears necessary to protect the heart against I/R injury by preventing the severe ER stress associated with ischemia and reperfusion. This concept is supported by the observation that inhibition of ER stress by tauroursodeoxycholic acid abrogated cardioprotection resulting from CyPD ablation or inhibition, while a moderate ER stress induced by tunicamycin reduced post-ischemic reperfusion injury [136]. More recent work from the Ovize laboratory suggests that CyPD contributes to Ca^2 + trafficking between ER/SR and mitochondria by binding to a protein complex which also includes VDAC1, Grp75 and IP3R1 [137]. This complex appears to be located at the interface between the OMM and the ER, and may facilitate the transfer of Ca^2 + released from the ER into the mitochondrial matrix [138]. Interestingly, the CyPD–VDCA1–Grp75–IP3R1 interaction was facilitated by mitofusin 2, and increased under conditions of hypoxia and reoxygenation [137]. This complex may be extremely relevant because the genetic ablation or the pharmacological inhibition of any of its members decreased the occurrence of cell death induced by protocols of anoxia/reoxygenation or ischemia/reperfusion [137].

Whether this additional role for CyPD is relevant to PTP regulation in relation to Ca^2 + is an open question. PTP opening does not depend on VDAC [61,62], and mitochondrial Ca^2 + uptake occurs through MCU in a Δψ_m-dependent process [139,140]. PTP opening curtails mitochondrial Ca^2 + uptake both by disrupting Δψ_m and by providing a possible pathway for Ca^2 + efflux [32–35], consistent with the finding that CyPD ablation or inhibition leads to an increase of matrix [Ca^2 +] [34,129,141].

An obvious question is whether can CyPD also be found outside the matrix and in the ER/SR. This question applies both to CyPD interactions with the VDAC1–Grp75–IP3R1 complex, and to those with other proteins shown to bind CyPD mostly by immunoprecipitation such as Bcl-2 [142], p53 [134,135], GSK-3 [131] and the Hsp90–Trap-1 complex [133]. These findings would imply that a fraction of CyPD is localized at sites other than the mitochondrial matrix, or that under specific pathophysiological conditions CyPD translocates to other cellular sites. Until such mechanism is identified, chances remain that results obtained with detergent extraction and immunoprecipitation may not reflect the protein–protein interactions involving CyPD in situ.

It should also be mentioned that the effects of CyPD on proteins located outside mitochondria need not be only due to direct protein–protein interactions, but might also be related to functional modifications induced by CyPD deletion or inhibition, as well as by covalent changes in response to the activation of signaling pathways [143]. These include (de)phosphorylation [131], (de)acetylation [144], nitrosylation [82,145] and oxidation of Cys²⁰³, which inhibits the enzymatic activity of CyPD and plays a crucial role in favoring PTP opening. Importantly, expression of a CyPD^C203S mutant desensitized the PTP to both Ca^2 +- and H₂O₂ providing a mechanistic link between oxidative stress and PTP opening [82]. This wide array of CyPD modifications, which has largely been characterized in E. Murphy's laboratory, is likely to affect mitochondrial metabolism, eventually resulting in variations of responses to physiological and pathological stimuli. In keeping with this possibility ablation of CyPD causes significant changes in the level of mitochondrial proteins involved in intermediary metabolism [146] as well as in the mitochondrial acetylome [147]. Interestingly, increased acetylation of β-hydroxyacyl-CoA dehydrogenase resulted in a 50% decrease of its activity [147], which may explain the decreased fatty acid oxidation observed in Ppif^−/− hearts [129].

CsA inhibits all CyP isoforms, and this can be an important confounder in evaluating its mitochondrial effects. Indeed, inhibition of calcineurin, which is mediated by the complex of CsA with cytosolic CyPA [92,93], prevents dephosphorylation of the pro-fission protein Drp-1 at Ser⁶³⁷, thus hampering its mitochondrial translocation and therefore fragmentation, a cytoprotective mitochondrial effect of CsA that is not related to PTP inhibition [148,149]. Of note, genetic ablation or pharmacological inhibition of Drp1 has been shown to afford significant protection against cardiomyocyte injury caused by ceramide, hyperglycemia and post-ischemic reperfusion [150–152].

This effect of CsA may be counteracted by other processes in the context of myocardial ischemia. Thus, lack of Ser⁶³⁷ dephosphorylation due to calcineurin inhibition can be contrasted by the activating effect of Ser⁶¹⁶ phosphorylation [153], a process catalyzed by various protein kinases including PKCδ that is generally considered detrimental for the survival of ischemic cardiomyocytes [153]. The protective effect of CsA could be further limited by direct stimulation of HIF-1α Pro⁵⁶⁴ hydroxylation, which leads to increased removal of HIF-1α and may thus prevent adaptation to hypoxia, an effect originally described in glioma cells [154] that should now be investigated in cardiomyocytes.

8. Summary and perspectives

The discovery that the PTP forms from the F-ATP synthase is redirecting research toward the mechanisms that switch this key enzyme from an energy-conserving to an energy-dissipating device. These hold great promise to improve our understanding of the pathophysiological events that trigger the transition in heart diseases, and to set a logical frame for therapeutic strategies. CyPD remains a viable target, and we are confident that the current hurdles between preclinical studies and clinical application of PTP-inhibitory strategies will be overcome by more specific inhibitors of CyP isoforms [155], by drugs targeting the PTP at sites other than CyPD [156] and by their combinatorial use.

Conflict of interest

The authors declare no conflict of interest.