Aconitase post-translational modification as a key in linkage between Krebs cycle, iron homeostasis, redox signaling, and metabolism of reactive oxygen species

Abstract

Aconitase, an enzyme possessing an iron–sulfur cluster that is sensitive to oxidation, is involved in the regulation of cellular metabolism. There are two isoenzymes of aconitase (Aco) – mitochondrial (mAco) and cytosolic (cAco) ones. The primary role of mAdco is believed to be to control cellular ATP production via regulation of intermediate flux in the Krebs cycle. The cytosolic Aco in its reduced form operates as an enzyme, whereas in the oxidized form it is involved in the control of iron homeostasis as iron regulatory protein 1 (IRP1). Reactive oxygen species (ROS) play a central role in regulation of Aco functions. Catalytic Aco activity is regulated by reversible oxidation of [4Fe-4S]²⁺ cluster and cysteine residues, so redox-dependent posttranslational modifications (PTMs) have gained increasing consideration as regards possible regulatory effects. These include modifications of cysteine residues by oxidation, nitrosylation and thiolation, as well as Tyr nitration and oxidation of Lys residues to carbonyls. Redox-independent PTMs such as phosphorylation and transamination also have been described. In the presence of a sustained ROS flux, redox-dependent PTMs may lead to enzyme damage and cell stress by impaired energy and iron metabolism. Aconitase has been identified as a protein that undergoes oxidative modification and inactivation in aging and certain oxidative stress-related disorders. Here we describe possible mechanisms of involvement of the two aconitase isoforms, cAco and mAco, in the control of cell metabolism and iron homeostasis, balancing the regulatory, and damaging effects of ROS.

Introduction

The tricarboxylic acid (TCA) cycle, also called the Krebs or citric acid cycle, is a central pathway of metabolism. It oxidizes acetyl coenzyme A produced by the catabolism of carbohydrates, amino acids, and fatty acids. Pyruvate dehydrogenase complex provides a link between glycolysis and the TCA. The reducing equivalents in the form of NADH and FADH₂ produced in the TCA, comprise the first step of oxygen-dependent ATP biosynthesis. The intermediates of the TCA cycle also are used in biosynthetic processes, primarily in the production of glucose, heme, lipids, amino acids, etc. The TCA cycle enzymes are located in the mitochondrial matrix, organized in a supramolecular complex that interacts with mitochondrial membranes and provides reducing power for the mitochondrial respiratory chain.¹

Many of the TCA cycle enzymes – aconitase (Aco), isocitrate dehydrogenase (IDH), fumarase, malate dehydrogenase – have cytosolic counterparts that participate in diverse extramitochondrial processes. In addition to their roles in the TCA, most Krebs cycle enzymes perform additional functions.² For example, Aco, α-ketoglutarate dehydrogenase, succinyl CoA synthase, and IDH are associated with the mitochondrial nucleoid, and may be involved in stabilization of the mitochondrial DNA, while IDH, lipoamide dehydrogenase, Aco, succinate dehydrogenase, and fumarase may be associated with mitochondrial mRNA translation, the oxidative stress response, iron metabolism, or tumor suppression.

Aconitase (aconitate hydratase; citrate hydrolyase; EC 4.2.1.3) catalyzes the reversible stereo-specific isomerization of citrate to isocitrate via cis-aconitate as an intermediate in a non-redox-active manner (Fig. 1). Aconitase belongs to the family of iron–sulfur-containing dehydratases. This enzyme undergoes reversible, citrate-dependent inactivation induced by oxidants, which cause disassembly of the iron–sulfur cluster.^3–6 Aconitase activity is widely used as a biomarker for oxidative stress and has been suggested to serve as an intramitochondrial sensor of redox status.^7–9 Numerous age-related and degenerative mitochondrial disorders are associated with elevated levels of reactive species with subsequent decline of Aco activity.¹⁰

Figure 1.

Aconitase functions within the cell. Cells possess two aconitase isoforms: cytosolic (cAconitase, ACO1) and mitochondrial (mAconitase, ACO2) ones performing both similar and different functions. The cytosolic one regulates an iron homeostasis as iron regulatory protein 1 (IRP1) when oxidized by reactive oxygen/nitrogen species (ROS/RNS). Mitochondrial aconitase converts citrate to isocitrate in the tricarboxylic acid cycle (TCA) and serves as a TCA cycle regulatory enzyme sensitive to oxidation. The higher respiration rate needed to increase ATP synthesis results in stimulation of ROS generation by the electron transport chain (ETC), primarily in the form of superoxide anion (O₂^•−). Further O₂^•− can react with nitric oxide (^•NO) yielding peroxinitrite (OONO⁻) or be converted to hydrogen peroxide (H₂O₂) either spontaneously, or enzymatically by superoxide dismutase (SOD). ROS/RNS interacting with aconitase decrease its activity and thereby slow down glucose consumption, respiration and ATP biosynthesis. The routes shown demonstrate interplay between regulation of cellular metabolism by aconitase and ROS/RNS.

Animals possess two Aco isoforms – cytoplasmic one (cAco) encoded by ACO1 gene and mitochondrial aconitase (mAco) encoded by ACO2 gene. The cytosolic Aco is also attributed to the iron-regulatory protein IRP1 and thus possesses dual functions. In its reduced form cAco acts as an enzyme, whereas in the oxidized form it can bind iron-responsive elements in stem-loop structures of specific mRNA, modulating biosynthesis of corresponding proteins involved in iron homeostasis.¹¹ Aconitase activity shows redox-dependent regulation, with involvement of the [Fe-S] cluster in redox modulation and redox-dependent posttranslational modification (PTM) of critical amino acid residues in the apoenzyme as discussed below.

The evidence that Aco is targeted by redox-dependent PTMs, both in the cytosol and mitochondria, suggests a key role for this protein in regulation of cellular stress-induced responses. Physiologic ROS fluxes from respiratory chain components of mitochondria can lead to production of minimally modified Aco forms, particularly of mAco, which may be a part of the complex series of responses identified within the term ‘redox signaling’. Higher ROS levels reached during stress, due to mitochondrial damage and NADPH-oxidase activation, are expected to cause more sustained PTM to Aco with functional impairment and subcellular redistribution of the proteins from mitochondria to the cytosol (reviewed in^12,13). Aconitase is one of the multitude of protein targets of oxygen and nitrogen reactive species so far identified in the literature (systematic reviews are presented in previous monographic works^14,15). Few of these protein targets, however, have demonstrated to produce functional effects relevant to biological and medical problems.

This minireview aims to analyse the knowledge accumulated to date on the potential role of aconitase as a link between regulation of Krebs cycle intermediates and iron metabolism with processes involving ROS – signaling and cell component modification. Although the two Aco isoforms, namely cytoplasmic and mitochondrial, apparently can perform different functions, the principles of PTM are common and involve both the cubic [4Fe-4S]²⁺ prostetic cluster and different oxidatively modified amino acid residues of the apoenzyme. This minireview is strictly focused on the issues mentioned above and does not cover all aspects of cellular processes. Therefore, we deeply apology to researchers whose works have not been cited here.

Cytosolic aconitase: a link between ROS metabolism and iron homeostasis

Aconitase has been found in both cytosol and mitochondria. The functions of each isoform, however, are different.^16–18 The regulation and maintenance of iron homeostasis within the cell is supposed to be a primary function of cAco. Since free iron ions can serve as electron donors for reduction of hydrogen peroxide leading to the formation of hydroxyl radicals cAco forms a link between iron and ROS metabolism. The possible involvement of cAco in ROS metabolism is illustrated in Fig. 2. Reactive oxygen and nitrogen species react with Aco to release labile α-Fe ion from the [4Fe-4S]²⁺ cluster resulting in the formation of a [3Fe-4S]⁺ cluster. Such modification leads to a switch in the protein function. Cytosolic aconitase (ACO1) in its oxidized form acts as iron-regulatory protein 1 (IRP1),^19,20 which recognizes iron responsive elements (IRE) of mRNAs of transferrin receptor (TfR) and ferritin and in this way regulates their translation.^20,21 IRE1 which possesses cysteine residues may be further oxidized by ROS. The levels of TfR and ferritine mRNAs were reduced in murine B6 fibroblasts in response to treatment with the redox cycling drug menadione sodium bisulfite.²² The described mechanism seems to be involved in fine regulation of cellular iron metabolism. An increase in free iron ion concentration in the cell may enhance the protein oxidation by hydroxyl radical generated in the reaction between iron and hydrogen peroxide. Oxidized proteins may undergo proteolytic degradation or form protein aggregates that inhibit proteasomal function.²³ At the same time, protein unfolding resulting from oxidation increases the antioxidant capacity of the protein due to higher exposure to the medium of the side parts of amino acid residues, particularly cysteine. Proteolytic degradation of the proteins, however, may further increase the polypeptide antioxidant capacity by formation of free peptides and amino acids. Both, protein unfolding and proteolytic degradation may serve as adaptive responses toward oxidative/nitrosative stress to decrease ROS/RNS concentration. A decreased concentration of ROS/RNS may prevent further Aco oxidation and, moreover, turn the function from iron regulation to enzymatic activity as aconitase. Markedly, a specific aconitase portion in cytosol possessing IRP1 or cAco activity appears to be determined by the dynamic equilibrium between these states of the protein and depends on cellular redox and iron status and/or phosphorylation.²⁴ The role of aconitase in iron metabolism regulation is provided by reversible oxidation and the reduction of [3Fe-4S]⁺ cluster to [4Fe-4S]²⁺ is known to be promoted by frataxin.^25,26

Figure 2.

Cytosolic aconitase links iron metabolism, unfolding and proteolytic degradation of proteins with ROS metabolism. Oxidation of aconitase by ROS/RNS results in loss of labile iron, inactivating the enzyme and switching the function to regulation of iron metabolism as iron regulatory protein 1 (IRP1). IRP1 may be reactivated by frataxin that assists to iron ion incorporation into the aconitase molecule and restores the [3Fe-4S]⁺ cluster to [4Fe-4S]²⁺. Decreased iron concentration prevents protein oxidation by hydroxyl radicals generated in the Fenton reaction. Oxidation of amino acid residues in the proteins induces protein unfolding and reinforces degradation by proteases. Both, unfolded and digested proteins were shown to have higher antioxidant capacity compared with normally folded protein and in this way may decrease steady-state ROS concentration.

Mitochondrial aconitase: a link between intermediate flow in the Krebs cycle and ROS generation

Mitochondrial aconitase catalyzes the conversion of citrate to isocitrate and is among the enzymes in the TCA cycle that are sensitive to oxidation. Using Drosophila as a model the effects of manipulating mAco gene expression were recently investigated with knockout and knockdown techniques.²⁷ Aconitase knockout flies were homozygous lethal, indicating that Acon is essential for viability. RNA interference-generated Aco-knockdown flies exhibited a variety of phenotypes, such as reduced locomotor activity, a shortened lifespan, and increased cell death in the developing brain. The known pathways of mAco involvement in regulation of metabolic flow for ATP generation, mitochondrial ROS production, and functional senescence are shown in Fig. 3. Modification of mAco by ROS derived from respiration or added exogenously may result either in oxidation of the [4Fe-4S]²⁺ cluster or cysteine residues.⁶ It has been shown that mAco can undergo reversible citrate-dependent inactivation in response to oxidants. Frataxin interacts with Aco in a citrate-dependent fashion, decreasing the intensity of oxidant-induced inactivation, and converting inactive [3Fe-4S]⁺ Aco to the active [4Fe-4S]²⁺ form.²⁸ For this reason a role for frataxin as an iron chaperone protein that protects the mAco [4Fe-4S]²⁺ cluster from disassembly and promotes enzyme reactivation has been proposed.

Figure 3.

Aconitase controls metabolic flow and ROS production. Mitochondrial aconitase can be oxidatively inactivated by loss of labile iron or by oxidation of cysteine residues. Frataxin restores the oxidized [3Fe-4S]⁺ cluster while oxidized cysteine residue(s) might be reduced by respective reductases. Partial loss of aconitase activity slows the metabolic flow through the TCA cycle and subsequent ROS generation. Decreased cellular respiration results in a decline in ATP level and may lead to cellular and tissue functional disturbance or senescence.

Interestingly, mAco also is involved in regulation of iron homeostasis in mitochondria. In a Drosophila model of Parkinson's disease Esposito et al.²⁹ found that inactivation of the Aco [4Fe-4S] cluster acts downstream of mitochondrial kinase PINK1 in a pathway that affects mitochondrial morphology and that this inactivation is responsible for iron toxicity. Clearly, although in different ways, both cytosolic and mitochondrial Aco, regulate iron homeostasis via the [4Fe-4S] cluster.

A decrease of mAco activity slows the flow of TCA cycle intermediates and generation of NADH and FADH₂ that are used as electron carriers for oxygen reduction in the electron transport chain (ETC) coupled with oxidative phosphorylation. Inhibition of electron transport through the ETC decreases ATP production. The latter has been shown to be connected with functional decline in aged organisms.³⁰ Age-dependent functional decline in cellular and tissue functions are enforced by increased steady-state ROS/RNS levels and are related to the oxidation of biomolecules involved in energy generation and overall homeostasis.

Exogenous ROS and RNS may affect Aco activity in vivo. It has been shown that yeast Saccharomyces cerevisiae Aco was inactivated by cell treatment with the redox cycling drug menadione sodium bisulfite in the presence of carbonate anion.³¹ Menadione also affected the enzyme activity in vitro. Rat treatment with another redox-cycling xenobiotic, the herbicide paraquat, enhanced cortical and striatal superoxide anion levels in parallel with decreased mAco activity.³² Additional evidence of in vivo modulation of Aco activity by exogenous nitric oxide was obtained through yeast treatment with sodium nitroprusside and S-nitrosoglutathione that are widely used as donors of nitric oxide.^33,34 Oxidation and retention of mAco has been proposed to represent a mechanism of yeast replicative senescence.³⁵ In isolated cardiac mitochondria treated with H₂O₂, aconitase was rapidly inactivated and subsequently reactivated in citrate-dependent manner.⁶

General principles of aconitase activity modulation: redox-dependent and independent posttranslational modifications

Reversible oxidation of iron–sulfur clusters and cysteine residues as well as reversible phosphorylation are well-known mechanisms of posttranslational Aco activity regulation.³⁶ The iron–sulfur cluster of the enzyme is oxidized by many oxidants. Loss of labile iron can result from the attack by reactive oxygen and nitrogen species of both intra- and extracellular origin. The operation of mAco is known to be modulated by the mitochondrial matrix protein frataxin, which acts as an iron chaperone. Frataxin plays a prominent role in mitochondrial iron storage and promotes the maturation of cellular iron–sulfur-containing and heme-containing proteins.²⁸ The possible role of maintenance of the active mAco pool has been described also for IscU protein involved in transfer of [4Fe-4S]²⁺ clusters to apo-aconitase.^37,38 Recently Sheftel et al.³⁹ found that ISCA1 and ISCA2, which are related to S. cerevisiae Isa1 and Isa2, respectively, and IBA57 are specifically involved in the maturation of mitochondrial aconitase and other [4Fe-4S] proteins functioning late in the ISC assembly pathway.

Cysteine residues are readily oxidized with the formation of sulfenic, sulfinic, or sulfonic acids, as well as homo- and mixed disulfides.^40,41 Only sulfenic acid and disulfides are recognized widely to be reduced with reductants such as glutathione and cysteine in vivo, or in vitro by dithiotreitol.^23,42 However, reduction of the sulfinic acid residue by sulphiredoxin was described also,⁴³ although it is not clear if it may occur in vivo. It was shown that modification of cysteine at or near the active site of purified mAco with the sulfhydryl reactive compounds N-ethylmaleimide or phenacyl bromide inactivated the enzyme.⁴⁴ Besides sulfenic acid formation, mAco cysteine residue modifications also include S-glutathionylation, which may be important in modulation of Aco activity under oxidative and nitrosative stresses.⁴⁵ All these PTMs centered on cysteine residues can be appropriate candidates for reversible redox-dependent modulation of the enzyme activity in mitochondria. It should be added that mAco after mild oxidative modification can be recognized and degraded by Lon protease, whereas severe oxidation may result in Aco aggregation, which makes it a poor substrate for the protease.⁴⁶ ATP-stimulated Lon protease was proposed to be essential to defend cells under oxidative stress.⁴⁷ Degrading slightly oxidized proteins Lon protease can prevent the accumulation of heavily oxidized proteins. This accumulation of heavily oxidized proteins may lead to poisoning of the cell due to which operation of Lon protease plays protective role under oxidative stress. Interestingly, the activity of Lon protease is suppressed under oxidative stress and aging.¹⁰ Saraswathy and Rao³⁶ found that in the retina during early experimental autoimmune uveitis aconitase was in the oxidized form.

Aconitase has been observed to undergo oxidation by ROS or products of their interaction with cellular components, such as reactive products resulting from lipid peroxidation like malondialdehyde (MDA), 4-hydroxynonenal, and other carbonyl-containing compounds (for a systematic review of ROS chemistry and lipid peroxidation see⁴⁸). Recently Liu et al.⁴⁹ found that interaction of mitochondrial aconitase with 4-hydroxy-2-(E)-nonenal resulted not only in inactivation of the [4Fe-4S] cluster, but also cysteine residues were found to be critical targets. The formation of additional carbonyl groups in Aco was described for aged flies⁵⁰ and mouse kidney.⁵¹ The enzyme activity was decreased by MDA treatment of mitochondria from heart and skeletal muscle of mice of different ages,⁵² which confirms the fact that it can be modulated by ROS and/or metabolites derived from reactions between ROS and cellular components. The formation of additional carbonyl groups in Aco was identified also in rat brain mitochondria.⁵³ Mitochondrial dysfunction is a common pathogenic feature of neurodegenerative diseases such as Alzheimer's disease (AD), Parkinson's disease,^54–56 Friedrich's ataxia, and others.⁵⁷ In AD patients dysfunctional mitochondria and a defective energy metabolism were associated with suppressed enzymatic Aco activity,⁵⁸ which was identified as one of the protein targets of 4-hydroxynonenal reactivity in early AD.⁵⁹ Interestingly, redox imbalance induced by perturbation of the homeostasis of NADPH may be responsible for inactivation of aconitase. For example, deficiency in the activity of glucose 6-phosphate dehydrogenase (G6PD) resulted in lower aconitase activity in a mouse model of human G6PD deficiency.⁶⁰

Mitochondrial operation induces a variety of PTMs in Aco and other associated proteins, which have functional downstream consequences for processes such as apoptosis, autophagy, and plasticity.⁶¹ Leakage of ROS from the mitochondrial electron transport complexes may sustain apoptotic cell death or even necrosis by the signals produced as a consequence of a defective energy metabolism and activation of the mitochondrial pathway of apoptosis.^12,13 N-Formyl-kynurenine formation from oxidized Trp residues in proteins has been tentatively identified as a source of PTMs of vicinal residues in certain mitochondrial proteins, including Aco.⁶² This early evidence needs more biochemical substantiation, but may suggest a role for the Trp metabolism in the pathophysiology of mitochondrial aging and dysfunction. Tyrosine residues are key targets of ROS to form PTMs such as 3-nitro-Tyr (3NT) and di-Tyr, which may interfere with pathogenic relevance in chronic diseases (reviewed in^62,63). Their formation may interfere with functional and signaling aspects of targeted residues, for instance, hampering the phosphorylation of Tyr residues of signaling proteins.⁶⁴ Conditions of nitrosative stress investigated in vivo in a flavohemoglobin mutant of S. cerevisiae, have revealed that 3NT may form on the mitochondrial proteins Aco and IDH, suggesting specific targeting of the citric acid cycle by RNS.⁶⁵ Accordingly, during nitrosative stress of the fungal pathogen Cryptococcus neoformans, Aco and other metabolic genes were found to be up-regulated as well as classical stress genes such as thiol peroxidase and gluthatione reductase,⁶⁶ thus suggesting a role for NO-induced PTMs in the response that coordinates different groups of genes involved in repair and adaptation with energy supply genes.

Peroxynitrite, the most biologically relevant ^•NO derivative of the cell produced by its interaction with superoxide anion,⁴⁸ inhibits Aco in concentration-dependent way and the enzyme substrate citrate was found to partially prevent this inhibition.⁴⁵ Biochemical analysis revealed nitration of tyrosine residues 151 and 472 and oxidation to sulfonic acid of cysteine residues 126 and 385 in Aco. All these modified tyrosine and cysteine residues are adjacent to the binding site and Cys 385 is one of the three cysteine residues in Aco that binds to the [Fe-S] cluster. These findings clearly demonstrate that oxidation and nitration reactions are responsible for the formation of PTMs thus leading to conformational changes, modification of regulatory, and active sites. ROS-induced PTMs of mAco also include enzyme interaction with N-formyl-kynurenine from oxidized tryptophan residues detected in bovine heart mitochondria.⁶¹ Peroxynitrite and H₂O₂ were proposed to regulate iron metabolism of mammalian cells by activating the iron regulatory proteins 1 and 2.^67,68 This effect requires a suitable cell redox state and follows the release of iron from the [Fe-S] cluster of cAco, which loses its enzymatic activity and gains binding affinity toward IREs of mRNA of specific proteins associated with iron metabolism.²¹ These regulatory effects were shown to be associated with the formation of 3NT on IRP1⁶⁹ and also of critical Cys residues on both IRP1⁶⁸ and IRP2.⁷⁰

One more aspect related to Aco operation should be mentioned here. Although redox-independent PTMs were not discussed above, we suggest that they may affect redox-dependent PTMs. Therefore, this problem will be briefly highlighted in the next paragraphs.

Aconitase activity can be regulated by means of redox-independent PTMs that include enzymatic phosphorylation and transamidation. Phosphorylation by protein kinase C was also proposed as a mechanism of Aco activity regulation.^71,72 It was found to be responsible for the increase in activity of its mitochondrial isoform in the heart of diabetic rats.⁷³ The mitochondrial aconitase isolated from rat tissues was identified as a substrate of mitochondrial kinases belonging to the Src kinase family that catalyze the reversible phosphorylation of Tyr residues Y71, Y544, and Y665.⁷⁴ Although the exact role of this selective phosphorylation remains unknown, the regulatory effects of this process in energy and signaling function of mAco are supposed. For example, up-regulation of mAco phosphorylation together with that of other energy-related proteins during adaptation responses occurring in vivo in the aged muscle, suggests its involvement in the promotion of energy supply under conditions of stress and overall increased metabolic demand.⁷⁵

A specific mechanism of Aco activity modulation in the brain was defined in Huntington disease, namely transglutaminase-dependent enzyme protein modification and aggregation.¹⁶ It was proposed that increase of transglutaminase activity was the potential mechanism responsible for this pathology and mAco was shown to be a primary transglutamination target in injured brains of these patients.¹⁶ Transglutamination of the enzyme resulted in the formation of high-molecular mass aggregates that was associated with loss of enzymatic activity.

Conclusions and perspectives

Posttranslational modulation of Aco activity plays a crucial role in regulation of cellular metabolism and iron homeostasis. Reversible oxidation of iron sulfur cluster and cysteine residues, phosphorylation, and transamidation, were described as key mechanisms of the enzyme activity modulation. Most of these regulatory pathways were shown in vitro, and in vivo studies are scarce. Here we discussed the possible mechanisms of modulation of both the cytosolic and mitochondrial, Aco isoforms and possible relationships with cellular metabolism, homeostasis of iron and reactive oxygen/nitrogen species. Despite in vivo studies describing the modulation of protein activity in cardiac ischemia/reperfusion, and functional decline across the age and diseases, additional studies should be performed to better understand the role of Aco in regulation of cellular and tissue functions. Aconitase protection by certain pharmacological approaches such as the application of a natural pigment with antioxidant activity, curcumin, may be promising in prevention of specific disorders.⁷⁶ It seems that Aco may be a convenient sensor to evaluate the redox status of both the cytosolic and mitochondrial cell compartments, but the question of correct measurement of its activity and prevention of experimental artifacts will need especial attention because up to now this is virtually not resolved and extensively debated.

Acknowledgments

We apology to many researchers in the field whose works were not cited in this minireview due to limited space. The authors also would like to thank an anonymous referee for very professional analysis of the paper and giving valuable suggestions resulting in better and more clear presentation of the material.