Mitochondrial permeability transition pore: a snapshot of a therapeutic target

KEYWORDS:

• Adenine nucleotide translocator
• ATP synthase
• cancer
• ischemia reperfusion injury
• mitochondria
• mitochondrial protein import systems
• permeability transition pore
• SLC25 transporters

1. Introduction

The mitochondrial permeability transition (MPT) is defined as the loss of inner membrane a specific impermeability and transmembrane potential, due to the sustained opening of a large pore (the MPTP). The pore (also nicknamed Mitochondrial MegaChannel, MMC) has been characterized decades ago in studies with isolated mitochondria and – by our group – by single-channel patch-clamp on rat liver mitoplasts. Very similar activity (the Multi-Conductance Channel, MCC) was independently described by K.W. Kinnally’s group (revs [1,2]). The awakening of this megachannel is elicited by increases in matrix Ca²⁺ levels and is potentiated by other agents/processes, the major one being oxidative stress. The equivalent diameter of the pore can be as large as 3 nm but the channel exhibits a variety of sizes. In samples of isolated mitochondria, the molecular weight of permeating molecules increases as a function of the inducing Ca²⁺ concentration and of the time elapsed since induction. Partial removal of the OMM would instead tend to decrease it [3]. Prolonged opening is accompanied by depolarization, release of matrix components, further generation of reactive oxygen species (ROS), morphological changes (swelling, fragmentation of the mitochondrial network), decrease in cellular ATP levels, activation of mitophagy and, eventually, cell death. MPTP opening has been proposed to be involved in ischemia/reperfusion injury (IRI), aging and aging-associated neurodegenerative diseases, chronic stress syndrome, traumatic brain injury, muscular dystrophy, diabetes, cirrhosis, pancreatitis, osteoporosis, and more. In addition, cancer cells deploy desensitization mechanisms which help cancer growth and dispersion, and which may oppose chemotherapy. Inhibiting or activating the MPTP is therefore an exceptionally relevant pharmacological goal. As such, it has received much attention, but so far successes in pre-clinical models have not been accompanied by a proportionate impact in clinical practice. The effort continues. The MPTP is also believed to have physiological and protective roles under normal circumstances, by opening transiently to act as a ‘Ca²⁺ release valve,’ preventing mitochondrial Ca²⁺ overload, and tuning mitochondrial function [4,5]. This function appears to be particularly important in cardiac myocytes.

A key area of research is the molecular nature of the MPTP, which has been investigated for more than 40 years and is still a matter of debate [6,7]. We will briefly recapitulate and comment on the current scenario brought about by recent developments in this field.

2. The molecular identity of mitochondrial permeability transition pore – an evolving story

The participation of Cyclophilin D (CypD), a mitochondrial peptidyl-prolyl cis/trans isomerase (PPIase), is widely considered a defining feature of the MPT, since both are inhibited by Cyclosporin A (CSA). Cyp D is, however, only a facilitator of the MPT, which can take place also in its absence or in the presence of CSA (although at higher [Ca²⁺]). For a long time, the major candidate for the role of pore former was the adenine nucleotide translocator (ANT). The candidacy of the ANT has its roots in the effects of two specific ANT inhibitors, atractylate (an inducer of the MPT) and bongkrekate (an inhibitor) (rev [8]). The ANT was demoted to the role of regulator when it was discovered that mitochondria from mice KO for two of the three ANT genes could still undergo the MPT (although, again, it required higher [Ca²⁺]) [9]. More recently, however, Molkentin and coworkers reported that resistance to Ca²⁺ became even stronger if the third ANT gene was also deleted, and complete if the CypD gene was knocked out as well [10]. Electrophysiological studies have confirmed previous reports that the ANT can form channels – although not as large as the ‘canonical’ megachannel [11,12].

A decade ago, Bernardi’s group presented evidence of a CSA-sensitive interaction of CypD with the oligomycin sensitivity conferring subunit (OSCP) of the F_OF₁ ATP synthase [13]. This introduced the possibility that Complex V might actually form the MPTP. One model now proposes that the MPTP may develop at the contact surface of monomers within the ATP synthase dimer [13], with Ca²⁺ acting on the ß subunit. The alternative model envisions the c-ring of the F_O sector as forming the pore [14]. The F₁ sector has been proposed to act as a gate of the c-ring [15]. The engagement of both the F₁ and F_O modules provides flexibility, much needed to explain all the properties of the pore. This model, while certainly appealing, needs confirmation. Another possibility is that the c subunit may form a large pore upon adopting a misfolded, amyloid-like fibrillar conformation [16]. Spoiling the picture, Walker’s group reported that various subunits of Complex V – the OSCP and subunit c among them – could be deleted, one at a time or in combination, even to the point of destroying the complex, without preventing mitochondria from undergoing a Ca²⁺-induced, CSA-sensitive permeability transition [17].

It is at any rate difficult to fit all data within a model based on only one pore. The idea that there may actually be more than one MPTP (e.g. [18]) has now taken hold. Most groups in the field agree that while the F_OF₁-ATPase is a major character on the MPTP scene, the ANT is a valid supporting actor. An executive summary might state that a smaller pore is formed by the ANT and a bigger one by the ATPase [2,5–8,11].

Using advanced microscopy techniques and whole cells treated with ferutin (a Ca²⁺ ionophore), Pavlov’s group has recently obtained evidence that both the ANT and at least the peripheral stalk of the F_OF₁ ATPase are actually required for a wide MPTP to form [19]. The study confirms that molecular interactions, presumably within the ‘ATP synthasome,’ are a key mechanistic aspect of the MPT. It also points to the largely unexplored possibility that the nature of the experimental biological material (purified proteins, isolated mitochondria, mitoplasts, whole cells, in vivo models) and of the applied experimental stimuli ([Ca²⁺], oxidizing agents, thiol reagents, small activating molecules) may influence the type of MPTP arising, and its properties.

3. Expert opinion

Once the idea that multiple proteins and mechanisms may underlie the MPT has entered mainstream thinking, one wonders whether other plausible MPTP candidates may exist. The answer is obviously yes.

The first that come to mind are other translocators of the Inner Mitochondrial Membrane (IMM), belonging, like the ANT, to the SLC25 family. In humans, this group counts 53 members with clear sequence and structure similarities [20]. These transporters function by alternating between states with the substrate-binding site accessible either from the cytoplasmic (c-state) or matrix (m-state) side. All members of the family have three homologous repeat domains, each with both C- and N-termini on the intermembrane side and two transmembrane segments. The latter are linked (on the matrix side) via a conserved signature motif, which includes a proline. After the paper by Kokoszka et al. [9] had weakened the position of the ANT as molecular substrate of the MPTP, Halestrap’s group proposed that the Pi carrier (PiC) might be a potential core component of the MPTP, capable of substituting the ANT [21]. The genetic deletion approach [22] confirmed an interaction of CypD and PiC but showed that the MPT could take place also in the absence of PiC, although it was blunted. This result is in line with all other cases in which this type of experiment has been carried out – for the Voltage Dependent Anion Channel (VDAC), the ANT, various subunits of F_OF₁, the Translocator Protein (TSPO) – and suggests that indeed multiple alternative mechanisms may underpin the MPT and result in functional redundancy.

Other serious candidates are the mitochondrial protein import systems. We have already mentioned this possibility in previous papers [1,18], but it has so far always been overshadowed by other appealing and well-supported candidates. The TIM17-TIM23 (rev [23]) and TIM22 (e.g. [24]) IMM import complexes have long been thought to have a twin-pore architecture. While this view has been challenged by recent structural studies, the patterns of their electrophysiological activity upon reconstitution in planar bilayers clearly resemble those of the MMC in mitochondria (compare, for example, the traces shown in [24] and [18]). Furthermore, there are strong similarities between the MMC of our patch-clamp studies and the MCC (e.g. [2,25]). The MCC has been identified as TIM23, mainly on the basis of the observed effects of mitochondrial import presequences on channel activity (e.g. [26]). Like the MMC, the pore has been reported to be activated by Ca²⁺ and inhibited by low pH. In yeast, a matrix cyclophilin accelerates the refolding of imported proteins, a process slowed down by CSA [27]. TIM22 activity could also be observed in mitoplasts after ‘awakening’ by specific signal peptides [28] and showed much of the same behavior as the MCC/TIM23, outer membrane TOM40 and MMC channels. More research needs to be done to determine to what extent SLC25 and TIM proteins may account for MPT occurrence. These systems have been largely characterized in yeast, hardly the best organism to study the MPT. Ad hoc work on the mammalian system would be demanding, but it may be very rewarding.

So far, attempts to intervene on MTP-related pathologies, focusing mainly on CypD, the F_OF₁ ATP synthase and, in the case of IRI, on ‘conditioning’ procedures, have had limited success. A major reason may simply be the redundancy of MPT machineries. Realizing that the mitochondrial killer may be a gang of accomplices rather than a loner may well advise pharmacologists to seek an approach acting at the level of induction rather than at that of pore formation. The major MPT-inducing factors are mitochondrial Ca²⁺ and ROS. Therefore, in the case of reperfusion damage, for example, an attractive target may be the mitochondrial Ca²⁺ uniporter complex (MCU), which transports the ion into the mitochondrial matrix. Several MCU inhibitors (as well as stimulators) have already been identified (revs [29]). At reperfusion, ROS are largely produced via oxidation by succinate dehydrogenase (SDH) of succinate accumulated during ischemia [30]. The use of rapidly hydrolyzing ester prodrugs of malonate, a competitive inhibitor of SDH, may thus be worth pursuing in this context [31]. Approaches obviously need to be custom-designed according to the specific therapeutic goal. At the other end of the spectrum, exploiting the MPT against cancer is not a new idea, but the time may have now come for renewed emphasis. In principle, in this case a proliferation of possible targets ought to be a favorable development: activating any one of them should cause a mitochondrial crisis and the death of the unwanted cell. Here, too, approaches featuring Ca²⁺ fluxes and ROS generation may be in the forefront. The redox stress strategy currently seems promising: neoplastic cells maintain intrinsically high ROS levels and are therefore especially vulnerable to additional redox stress.

Thus, paradoxically, a more complete definition of the proteins generating MPTPs may shift the focus of translational research towards a broader approach.

Declaration of interests

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript.

Reviewer disclosures

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Additional information

Funding

The authors are supported by the Italian Association for Cancer Research (AIRC grant n. 27271 to I Szabó, and fellowship n. 26584 to S Parrasia), by the Worldwide Cancer Research fund (WWCR grant n. 22-0348, to I Szabó), by PNRR (Spoke 2, to I Szabo), by Telethon (grant n. GMR23T2021, to I Szabo), by the Italian Multiple Sclerosis Society Foundation (FISM grant n. 2022/R-Single/046, to I Szabo) and by the Italian Ministry for the University and Research (PRIN grant n. 2022JPLKHW, to L Biasutto).