A Walk in the Memory, from the First Functional Approach up to Its Regulatory Role of Mitochondrial Bioenergetic Flow in Health and Disease: Focus on the Adenine Nucleotide Translocator

Surprisingly, in this paper we showed that, although ADP is essential to completely protect mitochondrial ANT-1, its effect on mPTP opening is only partial, thus suggesting that ANT-1 involvement is not the primary event leading to mitochondrial permeabilization and then to cell death, consistent with [145]. Furthermore, the complete prevention of death exerted by ADP but the partial prevention of mPTP opening confirmed that death and mPTP are not strictly dependent on each other in apoptosis.

Shifting the focus towards the effect of ATR on viability, this selective ANT inhibitor prevents ADP protection, thus suggesting that the molecular mechanism responsible for the ADP protective effect is not unique. Since complete prevention of ADP protection occurred only when ATR and pyridoxalphosphate-6-azophenyl-2′,4′-disulfonic acid, a functionally selective antagonist of purine receptor, were added together, it was inevitable to assume that the ADP protective effect on cell death is mediated both by the activation of the AOX system, occurring in cells undergoing apoptosis treated with ADP, and by its direct interaction with the mitochondrial carrier that drives the ADP/ATP exchange. Nevertheless, it should also be considered that the activation of AOX enzymes results in depressed levels of ROS, which in turn reduces the ANT-1 activity, as in [145].

In the same paper we observed that extracellular ADP also protects ANT-1 from the toxic action of the two AD peptides, i.e., Aβ1–42 andNH2htau, further corroborating the molecular mechanism of neuroprotection by ADP. ADP’s capability to protect the carrier from toxic NH2htau can be easily explained when considering that the inhibition exerted by this fragment on ANT-1 is competitive, that is, it interacts with the catalytic site to which the substrate binds.

On the assumption that neuropathies are the result of neuronal apoptosis, the identification of compounds that, like ADP, are able to protect neurons against apoptosis is highly desirable.

Studies on several amyloid precursor proteins and tau transgenic mouse models suggested that a possible link between these two characteristic AD hallmarks might be an early mitochondrial dysfunction, particularly at synapses associated with increased oxidative stress [162].

Interestingly, we observed that high intracellular levels of pathogenetic NH2htau are able to interfere with the mitochondrial biology leading to a drop in ATP and eventually to neuronal death, as reported in [149,150]. This is a typical example where ANT is the target of a toxic compound produced in the cell affected by AD.

A few years earlier, Amadoro et al. [151,152] showed that high intracellular levels of tau N-terminal fragments lacking the first 25 amino acids evoke a potent neurotoxic effect and induce a necrotic type of cell death, as sustained by protracted stimulation of N-methyl-D-aspartate (NMDA) extrasynaptic receptors. The NH2-26–44 tau fragment was the minimal active moiety, which retained a marked necrotic effect. In a meticulous study, we showed that the NH2-26-44 tau fragment can impair OXPHOS, as investigated in vitro with homogenates containing intact coupled mitochondria, with this occurring as a result of ANT impairment. Such an impairment could account for the reduced availability of ATP already proposed as the cause of the excitotoxic NMDA receptor activity mediated death in Alzheimer’s disease. In this case, the reduced ATP availability in the cytosol could in turn cause glutamate release from the cell and then excitotoxicity [151,152].

The experimental approach used to address the role of NH2-26-44 fragments in neuron energy metabolism highlighted that both cytochrome oxidase and ANT are targets of NH2-26-44 tau fragments, but ANT is the unique mitochondrial target responsible for impairment of OXPHOS. The mechanism by which the NH2-26-44 tau fragment impairs ANT is at present a matter of speculation. Since Yang and Ksiezak-Reding [163] showed that as a consequence of glycation, paired helical filament-tau from AD and advanced glycation end-tau generate oxygen free radicals and since Cente et al. [164] reported that expression of a human truncated variant form of tau protein leads to the accumulation of ROS, the possibility that ANT impairment derives from ROS–protein interaction as well as from protein–protein interaction has been considered.

In this regard, several studies exist in which different aspects of putative interaction sites, such as hydrophobicity, residue propensities, size, shape, solvent accessibility, and residue pairing preferences, are examined [149]. As far as the interaction of tau peptides and ANT is concerned, a simple comparison of the peptides and the ANT isoform sequences, based on the six-transmembrane model (three-repeat domain structure) proposed by Klingenberg [165], does not reveal any element that could be predictive of a direct interaction, such as cysteine or proline residues [149].

Nonetheless, it must also be considered that since the NH2-26-44 tau fragment is a non-competitive inhibitor, it does not interact with the catalytic but rather with some other site of the enzyme, which could distort the enzyme’s structure, thus also affecting the catalytic binding site. In particular, we believed it was important to characterize the mechanism(s) regulating ANT function not only for the purpose of explaining the pathways contributing to necrosis but also to understand the pathogenesis of various taupathies that may result from the altered function of ANT.

All together the results by Amadoro et al. [151,152] proved that caspase(s) play/s a dual pivotal role in apoptosis not only by degradation of endogenous tau, thereby reducing the intracellular endogenous pool of full-length tau available for binding to microtubules, but also by generating the production of the neurotoxic NH2-26–44 tau fragment that impairs the mitochondrial ANT and then ATP cellular availability.

The knowledge that mitochondrial dysfunction, with a key role played by ANT impairment, is of central importance in AD opens a window for new therapeutic strategies aiming to preserve/ameliorate mitochondrial function, and represents an exciting challenge for biochemists.

Subsequently, we showed that the NH2-26-44 peptide tau fragment preferentially interacts with Aβ peptide(s) in human AD synapses in association with mitochondrial adenine nucleotide translocator-1 (ANT-1) and CyP-D [149]. The two peptides inhibit the ANT1-dependent ADP/ATP exchange in a noncompetitive and competitive manner, respectively, and together further aggravate the mitochondrial dysfunction by exacerbating the ANT-1 impairment and thereby potentiating and amplifying each other’s deleterious effects [166]. The present work provides additional insights on how tau dysfunction directly causes mitochondria impairment at AD synapses by binding ANT-1 alone or in cooperation with a peptide(s). Consistently, an aggravated mitochondrial impairment has been described in eight- and 12-month-old APP/PS/tau transgenic mice combining both plaques and tangles, compared with mice overexpressing tau or APP alone. In this context, the pathological convergence between tau and Ab at AD synaptic mitochondria may help to explain why the diseased A- or tau-modifying strategies have not singularly given promising results and suggest potential new pathway(s) and target(s) [153] for a combined, more efficient therapeutic intervention of early synaptic dysfunction in AD.

Next, we dissected the molecular mechanism by which NH2-htau and Aβ1-42 impair mitochondrial ANT1. Since it is known that ANT contains –SH groups, which are both essential for the catalytic activity and preferential targets for ROS attack, we investigated whether ANT1 inhibition is due to an interplay between these two Alzheimer peptides, ROS and ANT1 thiols [167].

To accomplish this purpose, we turned to mersalyl, a reversible alkylating agent of thiol groups that are oriented toward the external hydrophilic phase [168], to selectively block and protect the –SH groups of ANT-1 in a reversible manner. The ability of Aβ1-42 to induce mitochondrial ROS production and that of mersalyl to protect ANT-1 against truncated NH2htau, but not Aβ1-42, was observed.

Then, a molecular mechanism in which the pathological Aβ-NH2htau interplay on ANT1 in Alzheimer’s neurons involves the thiol redox state of ANT-1 and the Aβ1-42-induced ROS increase was proposed. This result represents an important innovation because it suggests the possibility of using various strategies to protect cells at the mitochondrial level by stabilizing or restoring mitochondrial function or by interfering with the energy metabolism providing a promising tool for treating or preventing AD.

Our finding was consistent with the paper by Atlante et al. [124]: Singlet oxygen, produced by irradiation of photosensitizer-loaded mitochondria, triggers permeability transition pore (PTP) opening, and thiol reagents, in particular mersalyl, prevent PTP opening when mitochondria are irradiated after the addition of mersalyl. This is in agreement with the observation of Atlante et al. [123], according to which ANT is a major target of ROS-dependent photodynamic action (see above, Section 4.2.1).

Paradoxically, Aβ1-42 protects the cell from the inhibitory effect of NH2htau on ANT-1. In this regard, we observed that ADP added to cultured neurons also reduced the tau effectiveness in inhibiting ANT-1, like Aβ1-42 does. It might be possible that binding ADP to the carrier could grant protection against thiol oxidation by (i) promoting alterations of the ADP/ATP carrier conformation that change the position of its thiol groups [165,167,169], rendering them not accessible to oxidation by ROS, or (ii) changing mitochondria from the orthodox to a condensed configuration, an alteration that may protect thiol groups of membrane proteins against oxidation. This aspect is in agreement with the observation that ROS, while readily reacting with and damaging vital cellular structures, among them lipids, DNA, and proteins [168], can also regulate protein function through the modification of specific thiol(s). In recent years, an increasing number of proteins have been identified that are not damaged by oxidative stress conditions but use ROS-mediated thiol modifications to specifically regulate their function [168].

In the transition from the nerve cell, in which apoptosis is responsible for the advancement of neurodegeneration and death in the most famous diseases of the brain, first of all Alzheimer’s, to the plant cell, programmed cell death (PCD) is considered an integral part of the life cycle, being a major component of normal development, preservation of tissue homeostasis, and elimination of damaged cells in response to various forms of abiotic and biotic stresses [170].

The pivotal involvement of mitochondria in the execution of PCD is also undiscussed in plants [171], with characteristics observed in mammalian systems. In an original study carried out by Valenti et al., [172], with the aim of investigating whether and how mitochondria undergo changes in plant PCD, non-photosynthetic Tobacco Bright Yellow 2 (TBY-2) cells were used. In coupled mitochondria isolated from TBY-2 cells undergoing PCD as a result of heat shock, whether and how the ANT-dependent export of ATP synthesized via OXPHOS—and also the ATP production via ADK and the activity of the nucleoside diphosphate kinase (NDPK)—can change during the early phases of PCD in the TBY-2 cell line was investigated. A strong inhibition in the rate of ADP/ATP exchange via ANT with a drastic reduction in transport efficiency in spite of only a minor decrease in the amount of protein was found, suggesting that the overall effect is due essentially to a double impairment of ANT function, i.e., a minor decrease in the protein level and a strong decrease in the catalytic efficiency.

Mitochondrial dysfunction critically impairs nervous system development and is potentially involved in the pathogenesis of various neurodevelopmental disorders, including Down syndrome, the most common genetic cause of intellectual disability. Therefore, a punctual and functional study was made on certain individual steps involved in mitochondrial ATP production in human skin fibroblasts with chromosome 21 trisomy (DS-HSF), providing new insight into the molecular basis for mitochondrial dysfunction in DS [173].

In the experiments carried out in this study, already the reduced efficiency of externally added ADP in stimulating the rate of oxygen consumption in DS-HSF mitochondria was an indicator of altered OXPHOS capacity. It was therefore necessary to make use of spectroscopic measurements that allowed for the continuous monitoring of ATP synthesized by mitochondria and exported outside mitochondria arising from externally added ADP under conditions in which OXPHOS really occurs [104].

Once again, we availed ourselves of the ATP D.S. in order to investigate further the mitochondrial ATP-related energy production in DS-HSF. Since in mitochondria ATP efflux can occur with the contribution of both ATP synthase/ANT activities (in OXPHOS) and the intermembrane enzyme ADK (which itself, catalyzing the reversible high-energy phosphoryl transfer reaction between adenine nucleotides, can provide ATP from externally added ADP), in this case the authors also used experimental conditions designed to monitor the two pathways separately.

Surely one of the advantages of the ATP detecting method is being able to differentiate the origin of the newly synthesized ATP. In order to selectively measure ATP synthesized via OXPHOS, samples were treated with Ap5A to inhibit ADK [105]. When ADK-related ATP synthesis was selectively monitored, mitochondria were treated with CAT plus OLIGO to block OXPHOS.

DS cells showed an impairment of OXPHOS capacity involving ATP synthase, ANT, and ADK deficit, events that might be ascribed to post-translational modification: the deficit of ANT as well as of ADK activity in DS fibroblasts occurred despite both proteins being not only not differentially expressed, but also up-regulated at the protein level. Interestingly, exposure of DS-HSF to dibutyryl-cAMP, a permanent derivative of cAMP, stimulated ANT, ADK, and ATP synthase activities, whereas H89, a specific PKA (protein kinase A) inhibitor, suppressed this cAMP-dependent activation, indicating an involvement of the cAMP/PKA-mediated signaling pathway in the ATP synthase, ANT, and ADK deficit [173].

In a later study, the same authors showed that both epigallocatechin-3-gallate and resveratrol, natural polyphenols, reverse the severe impairment of mitochondrial bioenergetics and biogenesis in neural progenitor cells isolated from the hippocampus of Ts65Dn mice, a widely used model of DS that recapitulates many major brain structural and functional phenotypes of the syndrome, including impaired hippocampal neurogenesis [174]. The rescuing of the in vitro impaired neurogenesis, likely linked to the activation of the PGC-1α/Sirt1/AMPK axis, suggested that both EGCG and resveratrol exert a potential beneficial action for treatment in DS.

Recently, in contrast to what is the hottest research topic in cystic fibrosis (CF) disease, essentially focused on the cystic fibrosis transmembrane regulator (CFTR) protein [175,176], we first launched into a study essentially scouting mitochondrial bioenergetics, which was extremely necessary and essential to provide the groundwork for future research aimed at understanding the molecular mechanisms responsible for the involvement of mitochondria in CF and to identify the proteins primarily responsible for the F508del-CFTR-dependent mitochondrial impairment and thus reveal potential novel targets for CF therapy.

Hence, in the context of the research called ‘”Relationship between mitochondria and F508del-CFTR in Cystic Fibrosis”—supported by the Italian Cystic Fibrosis Research Foundation FFC#1/2015 Project to A.A.—with the aim of studying the role of mitochondria related to the deficiency of F508del-CFTR function, i.e., loss of phenylalanine residue at position 508, responsible for defective transport of chloride ions across airway epithelial tissues [177,178,179,180], we characterized mitochondrial function, in particular as it regards the steps of OXPHOS and ROS production, in airway cells either homozygous for the F508del-CFTR allele or stably expressing wt-CFTR.

We approached this issue essentially by using cell homogenate as a model system, as it contains intact and coupled mitochondria and provides a “physiological” environment for the investigation of the participation of mitochondria in cellular functions, thus allowing for simultaneous monitoring of the closely interlinked cellular and mitochondrial machineries. All the steps leading to mitochondrial ATP synthesis and to its export into the cytoplasm via ANT were examined one by one [181]. Consistently, an important note is that as ANT-dependent ATP availability in the cytosol regulates CFTR channel activity gating [182], it follows that the activity of the ADP/ATP translocator, in concert with cytosolic enzymes that consume ATP, is really critical in CF by playing a decisive role in the maintenance of cytosolic ATP levels.

Once again, also in this study, the procedure [104,145] that allows for the continuous monitoring of ATP efflux from mitochondria in the cell homogenate incubated with ADP appeared fit for the purpose of investigating ANT activity once we made sure that the rate of ATP efflux mirrored the rate of ADP/ATP exchange, which was verified by applying control flux analysis, as in [145]. We found that, together with other steps of OXPHOS, ANT-dependent ADP/ATP exchange was impaired in CF cells and, importantly, treatment of CF cells with the small molecules VX-809 and 4,6,4′-trimethylangelicin, which act as “correctors” for F508del CFTR by rescuing the F508del CFTR-dependent chloride secretion, significantly improved the mitochondrial parameters towards values found in the airway cells expressing wt-CFTR.