Redox control of enzymatic functions: The electronics of life's circuitry

Authors

  • Marcelo G. Bonini,

    Corresponding author
    1. Department of Medicine, College of Medicine, University of Illinois at Chicago, Chicago, IL, USA
    2. Department of Pharmacology, College of Medicine, University of Illinois at Chicago, Chicago, IL, USA
    3. Department of Pathology, College of Medicine, University of Illinois at Chicago, Chicago, IL, USA
    4. Programa de Biociencias Aplicadas a Farmacia (PBF), Universidade Estadual de Maringa, Maringa, Parana, Brazil
    • Address correspondence to: Marcelo G. Bonini, Department of Medicine, University of Illinois at Chicago, UIC, 909 S. Wolcott Avenue, COMRB 1131, Chicago, IL 60612, USA. Tel: +312–355-5948; Fax: 312-413-2948. E-mail: mbonini@uic.edu

    Search for more papers by this author
  • Marcia E. L. Consolaro,

    1. Programa de Biociencias Aplicadas a Farmacia (PBF), Universidade Estadual de Maringa, Maringa, Parana, Brazil
    Search for more papers by this author
  • Peter C. Hart,

    1. Department of Medicine, College of Medicine, University of Illinois at Chicago, Chicago, IL, USA
    2. Department of Pathology, College of Medicine, University of Illinois at Chicago, Chicago, IL, USA
    Search for more papers by this author
  • Mao Mao,

    1. Department of Medicine, College of Medicine, University of Illinois at Chicago, Chicago, IL, USA
    2. Department of Pharmacology, College of Medicine, University of Illinois at Chicago, Chicago, IL, USA
    3. Department of Pathology, College of Medicine, University of Illinois at Chicago, Chicago, IL, USA
    Search for more papers by this author
  • Andre Luelsdorf Pimenta de Abreu,

    1. Department of Medicine, College of Medicine, University of Illinois at Chicago, Chicago, IL, USA
    2. Department of Pharmacology, College of Medicine, University of Illinois at Chicago, Chicago, IL, USA
    3. Department of Pathology, College of Medicine, University of Illinois at Chicago, Chicago, IL, USA
    4. Programa de Biociencias Aplicadas a Farmacia (PBF), Universidade Estadual de Maringa, Maringa, Parana, Brazil
    Search for more papers by this author
  • Alyssa M. Master

    1. Department of Medicine, College of Medicine, University of Illinois at Chicago, Chicago, IL, USA
    2. Department of Pharmacology, College of Medicine, University of Illinois at Chicago, Chicago, IL, USA
    3. Department of Pathology, College of Medicine, University of Illinois at Chicago, Chicago, IL, USA
    Search for more papers by this author

Abstract

The field of redox biology has changed tremendously over the past 20 years. Formerly regarded as bi-products of the aerobic metabolism exclusively involved in tissue damage, reactive oxygen species (ROS) are now recognized as active participants of cell signaling events in health and in disease. In this sense, ROS and the more recently defined reactive nitrogen species (RNS) are, just like hormones and second messengers, acting as fundamental orchestrators of cell signaling pathways. The chemical modification of enzymes by ROS and RNS (that result in functional enzymatic alterations) accounts for a considerable fraction of the transient and persistent perturbations imposed by variations in oxidant levels. Upregulation of ROS and RNS in response to stress is a common cellular response that foments adaptation to a variety of physiologic alterations (hypoxia, hyperoxia, starvation, and cytokine production). Frequently, these are beneficial and increase the organisms' resistance against subsequent acute stress (preconditioning). Differently, the sustained ROS/RNS-dependent rerouting of signaling produces irreversible alterations in cellular functioning, often leading to pathogenic events. Thus, the duration and reversibility of protein oxidations define whether complex organisms remain “electronically” healthy. Among the 20 essential amino acids, four are particularly susceptible to oxidation: cysteine, methionine, tyrosine, and tryptophan. Here, we will critically review the mechanisms, implications, and repair systems involved in the redox modifications of these residues in proteins while analyzing well-characterized prototypic examples. Occasionally, we will discuss potential consequences of amino acid oxidation and speculate on the biologic necessity for such events in the context of adaptative redox signaling. © 2014 IUBMB Life, 66(3):167–181, 2014

Introduction

Plentiful oxygen in the terrestrial atmosphere has allowed for the evolution of diverse aerobic life forms but has imposed biological oxidations as a consequence of living on Earth. Fortunately, the electronic structure of the oxygen molecule limits its reactivity towards most substrates [1] because considerable activation energy is required to produce the reactive singlet state upon which a molecular orbital is vacated to receive an electron pair from diamagnetic substrates (the vast majority) (Fig. 1) (2). The singlet state, therefore, is the one that actually engages in most oxidations. In the mammalian cell, sophisticated mechanisms of enzymatic-assisted glucose breakdown to CO2 and water have evolved to optimize energy production and bypass the necessity to mobilize enormous amounts of activation energy to initiate oxygen-dependent reactions. Undoubtedly, such meticulous and intricate network is maintained functional by highly sophisticated enzymatic systems such as cytochrome c oxidase (CcO), the end point of the mitochondrial electron transport chain. CcO evolved the ability to transfer 4 H+ and 4 electrons extracted from NADH (by complex I) or FADH2 (by complex II) to molecular oxygen in serial reactions that reduces oxygen to water coupled to the buildup of the mitochondrial electrochemical potential (Fig. 2). The transfer of 4H+ and 4 electrons to oxygen prevents partial reduction that results in the formation of reactive oxygen species (ROS) (O2•−, H2O2, and OH). Although highly efficient, it is estimated that the mitochondrial electron transfer chain misses the mark of 100% efficiency by losing approximately 0.1–1% of the electrons along the way [3, 4]. Many studies devoted to finding the site(s) of electron leakage indicated that coenzyme Q (CoQ) which transfer electrons between complexes I and III is a major site of electron loss and consequentially of oxygen partial reduction/ROS production [5-7]. In addition, reactive oxygen and nitrogen species are produced by numerous other systems independently of mitochondria [8-13]. Fortunately, efficient antioxidant systems (enzymatic and non-enzymatic) evolved to mitigate detrimental effects of excessive ROS/RNS production and to regulate the locality, intensity, and duration of ROS/RNS mediated signals [14-19]. Because mechanisms of operation of such systems have been meticulously reviewed by other authors [20-28], in this review, we will only mention some in the context of protein oxidation repair. Noteworthy, even though most of the oxidants produced in vivo are scavenged and detoxified by efficient mutually regulated systems, oxidized proteins have been detected in various physiologic conditions and in nearly all pathologic states where oxidation seems to be widespread and exacerbated [29-37]. Oxidation modifies the electronic structure, polarity, size, and organization of target residues and therefore imposes significant changes to the ability of enzymes to properly interact with substrates and cofactors in specific and effective manners. Loss of catalytic ability is common and in rare cases enhancement in catalysis or gain of function was observed [38-40]. In any case, oxidation deregulates enzymes by either blunting their catalytic abilities, impeding their interactions with modulators, or altering post-translational modifications on regulatory residues (e.g., phosphorylation and acetylation). Of the 20 essential amino acids, four are particularly susceptible to oxidative modification: cysteine (Cys), methionine (Met), tyrosine (Tyr) and tryptophan (Trp) (see Fig. 3 for a schematic representation). Oxidation of cysteines and methionines is repairable by specialized systems devoted to reduce cysteine, mixed dithiols, sulfenic acids, and methionine sulfoxide. Modifications of tyrosine and tryptophan are typically terminal and when occurring at the active site determine the end of an enzyme's catalytic life. Among the most frequently detected tyrosine oxidation products in proteins, 3-nitrotyrosine, 3-chlorotyrosine, and dityrosine have been identified. The mechanism and implications of tyrosine modifications for enzyme catalysis will be examined later. Tryptophan is also susceptible to oxidation. Typically, tryptophan oxidation leads to the accumulation of degraded products resulting from the opening of the indole ring such as N′-formyl-kynurenine and kynurenine. Recently, an antibody against N′-formyl-kynurenine was developed and used to detect photooxidative damage to proteins in complex systems [41, 42]. In the next sections of this review we will attempt to critically examine the significance of specific oxidative modifications to enzyme catalysis and the cellular homeostatic balance. We will also attempt to highlight the frequently overlooked concept that different ROS and RNS vary greatly in their capacity to modify specific residues and that the different modifications themselves have diverse functional consequences. In this sense, ROS and RNS are not all equal and actually are very much dissimilar, contrary to what the acronym that harbors all of them under a familial umbrella appears to suggest. We hope to be able to highlight how important it is to be specific in determining the identity of ROS/RNS produced under exact circumstances and how necessary the utilization of probes that are able to discriminate them is in order to advance our knowledge of processes regulated by ROS/RNS in biology.

Figure 1.

Schematic representation of molecular oxygen's electronic configurations. In its ground triplet state (bearing two unpaired electrons in different orbitals), the oxygen molecule cannot accept a pair of electrons at once from a general diamagnetic substrate (S). Upon activation which produces the singlet state (no unpaired electrons and a vacant orbital) the oxygen molecule can readily accept a pair of electrons from a general diamagnetic substrate producing an effective oxidation reaction. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Figure 2.

Schematic representation of step by step electron and proton transfer reactions catalyzed by CcO. This complex (IV in the mitochondrial electron transfer chain) catalyzes the transfer of four electrons and four protons to O2 reducing it completely to H2O. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Figure 3.

Summary of the major oxidative modifications affecting tyrosine, methionine, tryptophan, and cysteine residues in biological systems.

ROS and RNS: What is Different Among them and How this Impacts Protein Oxidations?

Each ROS and each RNS is unique as every combination of atoms has a fingerprint set of electronic, magnetic, chemical, and physical properties (for an overview consult Table 1). Even the same combination of atoms arranged to produce different electronic structures differ greatly in reactivity, as exemplified by comparing 1O2 and 3O2. Observing recently published work it is easy to perceive that the practical acronyms ROS and RNS, used widely and interchangeably with more specific definitions, have given to many the false impression that all ROS and RNS are alike but set apart sharing an oxygen atom in their structure (in the case of ROS) or a nitrogen atom (in the case of RNS) not much else is common. ROS vary in size, polarity, diffusibility, oxidizing potential, electrophilicity, and hydrophilicity/lipophilicity. Therefore, some ROS will access residues buried deep inside enzymes' solvent inaccessible cores, while others will only oxidize residues exposed on the surface. This is the case of NO2 and CO3•− produced simultaneously from the decomposition of peroxynitrite in the presence of CO2 [43, 44]. Although NO2 reaches the solvent unexposed core of proteins (e.g., albumin) therein producing oxidations, CO3•−, due to its anionic character will react exclusively with solvent exposed residues [45, 46], reviewed in ref. 46. Not only free radicals but also non-radical oxidants are extremely important and distinct in their capacity to oxidize biomolecules. Likewise, it is critical to distinguish between them as the not so subtle differences in reactivity (hence effects) are various. Not all ROS are oxidants, for instance, superoxide radical anion (O2•−) is actually reducing at most physiologically relevant pHs [47] and in many cases effects attributed to O2•− maybe due to H2O2, an unavoidable dismutation product, or to other oxidants generated from O2•−, for example, peroxynitrite or even CO4•− [48]. Because nature itself is specific in the cellular mechanisms it created to scavenge O2•−, H2O2, and so forth (which can be potentially manipulated for therapeutic purposes), the necessity to distinguish which among ROS and RNS is fairly evident. Reactive species also combine and react amongst themselves, neutralizing or potentiating their destructive potential (e.g., ref. 49–51). An important example is the reaction of NO with O2•− to produce peroxynitrite (ONOO), a highly reactive, extremely powerful oxidizing agent that acts by exerting one and two electron oxidations (for a review please see ref. 46). This reaction transforms two relatively harmless species NO and O2•− into a strong oxidizing agent ONOO/ONOOH. In complex systems, many reactive species live and die together, that is, they coexist and interact and thus very rarely will a single assay be sufficient to provide a reliable picture of the origins and consequences of specific biomolecule oxidation in living systems. The work of tracking specific effects back to particular effectors is laborious and tedious when it comes to investigating the origins of redox processes; however, new [52-58] and improved methods [59-64] have brought promising and effective techniques to the field in the past few years.

Table 1. Chemical-biological aspects of major ROS/RNS produced in living systems
Species Origin (most relevant biological sources)Half-lifeapKaReduction potential (V)Citation
  1. a

    Estimated from known rate constants with relevant targets in biological systems.

  2. b

    Augusto, O., Miyamoto, S. Oxygen Radicals and Related Species. In: Principles of Free Radical Biomedicine. Volume 1. Editors: K. Pantopoulos, H. M. Schipper.

Singlet oxygen1O2Photochemical reaction (dye-sensitized photooxidation); chemical reaction (HOCl with H2O2)65 µs in D203.5 µs in H20<1 µs in physiologic conditionsN/A+0.65[65-70]
Superoxide anionO2•−Leaking electrons from mitochondrial ETC; NADPH oxidase; Xanthine Oxidase; uncoupled NOS1 µs to 1 ms4.8Eo O2•−,2H+/H2O2 = +0.94Eo O2•−/O2 = −0.33[47, 71-74]b
Hydrogen peroxideH2O2Dismutation of superoxide; Glucose oxidase catalyzed oxidation of glucose; xanthine oxidase; aldehyde oxidase; monoamine oxidase10–100 µs11.6+1.77[75]b
Hydroxyl radical•OHdecomposition of hydroperoxides1 nsN/A+2.31b
Nitric oxideNOoxidation of l-arginine by nitric oxide synthase (NOS); NO2 in acidic environments100 ms to 1 SecN/AEo NO/3NO = −0.80[76-79]b
PeroxinitriteONOOdiffusion-controlled reaction (k > 6.7 × 109 M−1 S−1) between superoxide and nitric oxide1.9 Sec at pH 7.4 in phosphate buffer<10 µs to <1 msIn vivo6.5–6.8+2.31 (one electron via hydroxyl radical)+1.4 (two electron)[49, 80, 81]

Oxidation of Cysteines: Reactivity, Reversibility, Implications, and Consequences

Cysteines are possibly the most oxidation-prone constituents of proteins and enzymes. As any thiol, cysteines can be oxidized by one- and two- electron transfer mechanisms producing thiyl radicals [82, 83] and sulfenic acids [24, 84], respectively. Further reactions of thiyl radicals and sulfenates differ but occasionally, in an environment rich in thiols, thiyl radicals, and sulfenates react to produce disulfides [82, 85, 86]. Once formed, thiyl radicals engage in reactions with molecular oxygen and other thiols at rates close to the diffusion limit (Fig. 4) producing sulfenyl and disulfide radicals [82]. Disulfide radicals are highly reducing and easily donate the unpaired electron to electron acceptors, most often molecular oxygen, producing superoxide radicals. Sulfenic acids are reactive towards reduced thiols and in their presence readily produce dithiols and water. Thus, in either case, oxidation of thiols by either one or two electron mechanisms, in the physiologic cellular milieu, produces dithiols (frequently, mixed protein-S-S-glutathione adducts). Indeed, the finding of glutathiolated proteins associated to measurable effects in signaling spurred considerable discussion and controversy as to whether in certain cases the activating/inhibitory enzymatic modification involved thiyl radical intermediates or was directly mediated by the reaction of an oxidized thiol residue with glutathione, and therefore imposed by structural changes exerted by the glutathione moiety on the enzymatic domain(s) to which it binds. Distinguishing between both is not trivial and in the specific case of Ras-GTPase regulation by oxidants required detailed studies involving the direct detection of thiyl intermediates formed on Ras by immuno-spin trapping, NMR analysis of glutathiolated Ras (that showed negligible active site structural changes imposed by GSH binding), and measurements of GTPase activity and nucleotide exchange rates by native and modified Ras [87, 88]. These studies put in context highlight the fact that protein-S-SG adducts should be regarded cautiously as they are a common end products of intricate processes and may not necessarily be the cause of enzymatic activity changes when the modified cysteine residue is not directly involved in catalysis.

Figure 4.

One- and two- electron transfer oxidations of thiols. Reactions of thiols with free radicals (one electron oxidation pathway) and non-radical oxidants (two electron oxidation pathway) are represented as well as some of the reactions of derived thiyl radical (one electron oxidation pathway) and sulfenic acid (two electron oxidation pathway). [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Cysteines can be oxidized beyond the sulfenic acid level to sulfinic and sulfonic acids (Fig. 5A). In the case of enzymes that depend on key catalytic cysteines for function, oxidation beyond reversibility signifies the end of the enzyme's catalytic life. Nevertheless, sophisticated mechanisms of protection of enzymatic function have been demonstrated which lowers the propensity of reactive cysteines to undergo irreversible oxidation. One example is the formation of sulfenyl-amides between the peptide bond nitrogen and cysteine–sulfenic acids intramolecularly [89]. In the case of protein tyrosine phosphatase 1B (PTP1B), the sulfenyl-amide intramolecular adduct has been shown to form under conditions of oxidative stress and to prevent active site cysteine oxidation to higher sulfoxides. Also, it has been shown (in the same study) that sulfenyl-amide intermediates can be reduced by thiols regenerating fully functional PTP1B (Fig. 6). This finding supports the authors' claim that molecular mechanisms have evolved to protect active site cysteine from being oxidized beyond repair [89]. Under intense or persistent oxidative stress, however, cysteine residues become oxidized to higher sulfoxides, sulfinic ([BOND]SO2) and sulfonic ([BOND]SO3) acids. These higher sulfoxides were once considered to be terminal modifications repairable solely by the degradation of the affected protein and recycling of the unaffected amino acids. Work pioneered by Rhee and colleagues showed, however, that sulfiredoxins can reduce sulfinic acid intermediates back to cysteine at the expense of ATP [90-92]. Therefore, it is possible that in an energy consuming process [BOND]SO2 groups can be reduced under very specific physiologic conditions when oxidative stress is transient or localized. This is because generalized and/or persistent oxidative stress in many contexts dampens ATP production and imposes severe restrictions to the functioning of systems prolific in ATP generation such as mitochondria and the glycolytic pathway [93-95]. Supposedly under ATP shortage the efficiency of [BOND]SO2 reduction would be expected to be diminished. Differently than [BOND]SOH and [BOND]SO2, sulfonic acids ([BOND]SO3) represent a terminal and irreparable modification of cysteine that would lead to the permanent loss of enzyme function and possibly degradation. Interestingly, examples exist of cysteine oxidation to sulfonic acid that leads to the formation of functional modified enzymes. For instance, Tsa1 peroxiredoxin was demonstrated to have its chaperone activity enhanced fourfold while losing its peroxidase function when oxidized [90]. The pathophysiologic implications of these findings are undoubtedly intriguing but remain to be fully clarified in the context of mammalian cell signaling.

Figure 5.

Schematic representation of progressive oxidative modifications of (A) Cysteine and (B) Methionine.

Figure 6.

Schematic representation of mechanisms of “protection” of critical catalytic thiolate moieties on (A) PTEN and (B) PTP1B active sites. These intermediates are believed to be produced upon the oxidation of the catalytic thiol and are reversible. Therefore, they are proposed to serve as “built in” breakers of oxidative stress that impose restrictions to the irreversible oxidation of cysteines that are essential for catalytic function. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

The oxidation of active site cysteines affects enzymatic catalysis and many are the examples of proteins that are regulated by such mechanism [96, 97]. These include but are not limited to caspases [98], phosphatases [99-101], proteases [102, 103], and even transcription factors [104]. Indeed, it is very likely that most proteins and enzymes are susceptible to functional redox regulation by the oxidation of component thiol residues to some degree. Below, we will examine prototypic examples of enzymes known to be affected by cysteine oxidation while analyzing the consequences of such occurrence to cell signaling, function, and homeostasis.

Thiolate-Dependent Phosphatases: Critical Nodal Points of Redox Signaling in Health and Disease

Many phosphatases identified to date are dependent on a critical low pKa catalytically active thiolate moiety that when oxidized invariably leads to enzyme inactivation. Although cysteine and glutathione thiol pKa is in the alkaline range (Cys, pKa = 8.22; GSH, pKa = 9.42) [105], sophisticated structural motifs have evolved in protein phosphatases that lowered the active site cysteine pKa by several units placing most in the range of ∼4.5–5.5 [106]. Certainly, the low pKa of these residues, that is required for catalysis, also make them especially susceptible to oxidation as discussed above. Phosphorylation effected by kinases and counteracted by phosphatases represents a major mechanism of cell signaling regulation, a basic aspect of cellular signaling circuitries and an imperative for the orderly and integrated management of nearly all processes that occur in cells and tissues. Predictably, the redox inactivation of thiol phosphatases results in a shift of signaling towards the prevalence of unopposed kinase-activated reactions. Invariably, oxidative stress, either transient or persistent, has been associated with increases in kinase activity for the most part consequentially to thiol phosphatase inhibition [107-109]. The inability to restore phosphatase function with persistent prevalence of the phosphorylation reaction has been, in many instances, associated with the onset and progression of diseases [38, 110-113]. In this section, we will review two examples of phosphatase activity inhibition by catalytic cysteine oxidation. We chose to examine phosphatase and tensin homolog (PTEN) and PTP1B, two prototype phosphatases with prominent roles in signaling and fairly typical reactivity towards biological oxidants, albeit different mechanisms of protection against irreversible oxidation.

Phosphatase and Tensin Homolog Deleted on Chromosome 10 (PTEN)

PTEN is a dual lipid and protein phosphatase [114, 115]. The lipid phosphatase activity of PTEN is by far the most studied because of its regulation of 3,4,5-phosphatidyl-inositol triphosphate abundance (3,4,5-PIP3). By catalyzing 3,4,5-PIP3 conversion to 4,5-PIP2 PTEN counteracts phophatidyl inositol-3-kinase (PI3K), turning its signal off. PTEN is depleted in 30–40% of human cancers concomitant to marked increases in Akt activity [116, 117]. Therefore, PTEN is generally referred to as a tumor suppressor. In PTEN+/− mice, PTEN deficiency has been shown to increase propensity to tumor development at multiple organs at an unusually early age [118-121]. Homozygous PTEN-deficiency is embryonic lethal and represent a compelling example of the necessity of an orchestrated kinase/phosphatase balance for development and the maintenance of cellular homeostasis [122]. Oxidation, in addition to well-characterized mutations, impairs the capacity of PTEN to hydrolyze 3,4,5-PIP3 and this effect was pinpointed to the oxidation of the active site thiolate Cys124 (Fig. 6B). In fact, Cys124 of the PTEN catalytic site has a particularly low pKa that renders it stable in the especially reactive thiolate form critical for the hydrolysis of the inositol-phosphate bond [123]. However, the same chemical/structural particularities that enhance Cys124 of PTEN reactivity towards 3,4,5-PIP3 make it vulnerable to oxidation. Many studies have identified PTEN oxidation as causative of cell signaling alterations associated with oxidative stress [99, 124, 125]. In particular, PTEN Cys124 has been shown to be sensitive to H2O2 [99, 126] and peroxynitrite [127]. Upon oxidation, PTEN Cys124 rapidly reacts with Cys71, producing an intramolecular dithiol. It is likely that the formation of PTEN Cys124-Cys71 intramolecular disulfide exemplifies another protective mechanism that prevents further oxidation of the enzyme, thereby warranting the reconstitution of PTEN function upon dissipation of the oxidant gradient [126]. An intriguing particularity of PTEN is the easy distinction of the oxidized and reduced forms of the enzyme through non-reducing SDS-PAGE/Western blot [126]. In its closed conformation, the oxidized form of PTEN migrates faster through the acrylamide gel due to the limited frictional resistance of the more compact globular shape assumed by the oxidized form. Detection of oxidized PTEN is a reliable indication of reduced activity that was confirmed with dominant negative mutants of PTEN lacking functional Cys124.

Protein Tyrosine Phosphatase 1B

PTP1B is a cytosolic phosphatase best characterized as a negative regulator of insulin signaling and, therefore, a potential drug target for the treatment of diabetes. More recently, PTP1B has been proposed to have many other roles stemming from the identification of its direct or indirect regulation of kinases that serve as nodal points in cell signaling such as AMP-activated protein kinase AMPK [128], mitogen-activated protein kinase (MAPK) [129, 130], RhoA/Rac1 [131], and the insulin receptor itself [132]. PTP1B is sensitive to slight variations in the intracellular steady state concentration of biological oxidants and thus serves as a major integrator of redox signaling in the cell and mediator of adaptative responses to oxidative stress [133, 134]. PTP1B function depends on the catalytic cysteine (Cys215) residue and its chemical modification has been shown to ablate the enzyme's phosphatase function (reviewed in ref. 135). Because of the critical role PTP1B plays in integrating outside-in signaling through directly affecting receptor activity and downstream effector function, it is not surprising that mechanisms have evolved to protect the enzyme from irreversible oxidative inhibition. Diverse mechanisms have been proposed/demonstrated. One such mechanism involves the nitrosation of Cys215, a reversible modification that prevents further oxidation [136] (Fig. 6C). Such mechanism may be particularly important under conditions of increased NO production such as inflammation, iNOS-expressing cells and in the vascular endothelium where NO production is kept steady by a fine balance defined by signals that activate and repress nitric oxide synthase function. Another mechanism has been demonstrated that involves the formation of sulfenyl-amide intermediate between Cys215-SOH and the peptide bond nitrogen of Ser216 [89]. The elegant study by Salmeen and colleagues showed that the sulfenyl-amide bond is resistant to further oxidation and promptly reversed by thiols thereby constituting an efficient mechanism of protection of the active site under conditions of transient oxidative stress.

Methionine Oxidation

Oxidation of protein methionine (Met) residues is an important and often overlooked post-translational modification resulting from persistent oxidative stress [137-139]. Methionine oxidation can either be reversible (when the amino acid is oxidized to methionine sulfoxide) or irreversible (when oxidation is pushed further to generate methionine sulfone). The effects of methionine oxidation in most biological systems are still underappreciated, but it is becoming clearer that methionine sulfoxide formation may broadly impact protein function and enzymatic activity while ultimately leading to pathogenic signaling [137-139]. Notably, calcium/calmodulin-dependent protein kinase II (CaMKII) has been shown to be activated following methionine oxidation that renders the enzyme constitutively active thereby abrogating the dependence on calcium for activity regulation [139]. In the heart, CaMKII oxidation has been linked to cardiac failure and increased propensity to arrhythmias [140-142]. In breast cancer epithelial cells, CaMKII methionine oxidation was proposed to enhance glycolytic activity relieving metabolic barriers for cancer cell survival via activation of AMPK [143]. These examples from vastly distinct models indicate that methionine oxidations have broad and important implications for cell signaling. Methionine oxidation, however, can be reversed by methionine sulfoxide reductases (e.g., MsrA and MsrB) at the expense of thioredoxin and NADPH [144]. In fact, consistent with a role for methionine oxidation in tumor progression, it has recently been shown that MsrA is downregulated in biopsied tissue of human breast cancer in a stage-dependent manner [138]. In the same study, authors showed that repression of MsrA paralleled by increased H2O2 and protein oxidation led to increased invasiveness and tumor growth [138].

The importance of methionine redox status may also be extended to various other pathologies, including Alzheimer's disease [137] and Parkinson's disease [145-147]. In line with this notion, methionine oxidation of α-synuclein by H2O2 was shown to cause instability and aberrant conformational changes of the protein that likely contribute to development of Parkinson's disease; however, this modification of α-synuclein has been shown to be reversed by MsrA reduction of methionine residues indicating an important role for MsrA in the maintenance of α-synuclein structural integrity [145, 147]. Further, it has been demonstrated that the frequency of oxidized methionine on α-synuclein differentially affects its oligomeric stability and fibrillation kinetics which are both important determinants in Parkinson's disease pathogenesis [145, 146]. Be it through alteration of conformation or by directly impacting enzyme catalysis, methionine oxidation is emerging as an important reversible regulatory mechanism of cellular homeostasis, viability, and function.

Cellular Metabolism, NADPH, and the Reversibility of Thiol Oxidative Modifications

Except for some peroxiredoxins that can be directly reduced by ascorbate [148], most thiol and methionine oxidative modifications are ultimately reversed by systems whose functioning depends on NADPH (Fig. 7). These include the glutathione peroxidase/glutathione/glutathione reductase (GPx/GSH/GSR) and thioredoxin/thioredoxin reductase (Trx/TR) electron shuttles. The former maintains the intracellular reduced glutathione pool in large excess over oxidized glutathione, thereby allowing for the maintenance and rapid recovery of reduced active thiolates on proteins and enzymes. The latter is the primary recovery system of oxidized methionine and some thiol-based enzymes. Both systems are fed by electrons subtracted from NADPH produced (primarily) by the pentose phosphate pathway. Therefore, the very maintenance of reductive and reparative capacity of the intracellular milieu ultimately depends on glucose availability and utilization in the cells. In a recent report, Jeon and colleagues found that AMPK activation is a fundamental requirement for the maintenance of NADPH levels in cancer cells [149]. Interestingly, most cancer cells are under marked oxidative stress which apparently confers survival and proliferative advantages to these highly adapted cells [150-153]. Likewise, in light of this observation, a steady supply of large amounts of NADPH must be critically important for cancer cells to maintain essential redox sensitive signaling networks viable. In a way, glucose is the ultimate sink for biologic oxidants in the cell. Thus, glucose uptake is crucial for fast proliferating cancer cells that generate high levels of oxidants. This thought resonates well with recent observations related to the metabolic transformation of the cancer cell as tumors progress to advanced stages [154, 155] and the proposition that increased glucose levels and hyperinsulinemia promote tumorigenesis in diabetics by increasing insulin-dependent glucose uptake by tumor cells [156, 157]. Thus, it is possible that the metabolic state of the cell regulates many aspects of cell signaling sensitive to redox regulation by defining whether sufficient NADPH levels are maintained to absorb the impact of increasing steady state levels of oxidants.

Figure 7.

Summary of reductive pathways responsible for repairing oxidized methionine and thiol moieties in proteins. Note the essential role of NADPH produced mainly (but not exclusively) by the breakdown of glucose in the phosphate pentose pathway. Glutathione peroxidase (GPx); glutathione reductase (GSR); thioredoxin (Trx); and thioredoxin reductase (TrxR) are represented by abbreviations in the scheme. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Oxidative Modifications Tyrosines: Mechanisms, Implications for Signaling, and Significance

Oxidation of the phenolic moiety of tyrosines (Tyr) is an irreversible and generally detrimental modification for protein/enzyme structure and function. How profound the impact of tyrosine nitration, hydroxylation, halogenations, or cross-links will be certainly depends on whether a particular tyrosine residue is involved in direct regulation of signaling (by tyrosine phosphorylation) or compose specific structural motifs that are required for proper interactions between binding partners. Mechanisms of tyrosine modification are various but, in vivo, always depend on the production of a biological peroxide, nitric oxide, or reactive species (mainly HOCl, HOBr, and peroxynitrite). Below we present some of the work devoted to unraveling the mechanistic basis of tyrosine chemical modifications while discussing their relevance for enzymatic functions.

Protein Tyrosine Nitration

Tyrosine nitration results in the production of 3-nitrotyrosine which are typically eliminated by protein degradation. Tyrosines can be nitrated on the ortho position of the phenolic ring by distinct mechanisms involving the formation of a precursor tyrosyl radical. For many years, 3-nitrotyrosine was considered a biomarker of peroxynitrite formation in vivo. In fact, in the presence of CO2, peroxynitrite efficiently nitrates protein tyrosines [158, 159] via a concerted mechanism in which carbonate radicals, resulting from peroxynitrite reaction with CO2, oxidize tyrosine to a tyrosyl radical that efficiently recombines with nitrogen dioxide (reviewed in ref. 46) generating protein bound 3-nitro-tyrosine. Later work, nevertheless, showed that tyrosine nitration can be effected in vivo by heme peroxidases operating in the presence of hydrogen peroxide and nitrite [160, 161]. In this case, heme peroxidases using nitrite as a substrate produce both the tyrosyl radical intermediates and nitrogen dioxide for nitration. These posterior studies demonstrated that 3-nitrotyrosine can arise from different oxidative events that, however, obligatorily involve nitrogen dioxide production. Therefore, it is still safe to assume that 3-nitrotyrosine formation is a biological signature of oxidative stress. Apart from mechanisms that involve nitrogen dioxide, an alternative mechanism for the pathophysiologic production of 3-nitrotyrosine was proposed in which a tyrosyl radical recombines with NO to produce 3-nitrosotyrosine. Oxidation of the relatively unstable 3-nitrosotyrosine intermediate would then produce 3-nitrotyrosine [162-164].

Nitration of tyrosine introduces a relatively bulky nitro group in the vicinity of the phosphorylation site on the phenolic hydroxyl group, alters the polarity of the residue and the microenvironment where it is inserted thereby disrupting hydrogen bonding and van der Waals interactions that may be important for the maintenance of the protein structure or catalytic site integrity. Tyrosine nitration is, therefore, a disruptive modification that can (and often does) hinder protein function.

A fairly well-acknowledged example of loss of enzyme function produced by tyrosine nitration is MnSOD [165], a manganese metalloprotein involved in the governance of mitochondrial ROS metabolism. Nitration of Tyr34 on MnSOD has been shown to render MnSOD unreactive towards superoxide radical anion which normally is dismutated into O2 and H2O2 by MnSOD. This is because the nitro group on Tyr34 blocks the access of superoxide radicals to the manganese atom located at the end of an anion channel [166].

Another example regards the nitration of cytochrome c Tyr74 which produces conformational changes sufficiently profound to impact the ability of the protein to participate in electron transfer reactions [167] or trigger proapoptotic responses [168]. In a detailed study of the consequential effects of cytochrome c nitration, it was found that the electron transferring function of cytochrome c is replaced by gain of peroxidase function [169], certainly producing major impediments for the proper functioning of the electron transport chain, mitochondrial bioenergetics, and the cell.

Taken together, many studies devoted to the understanding of the mechanisms of tyrosine nitration have indicated that some residues are more vulnerable than others and that the microenvironment where the tyrosine residue is located (either within the protein or in specific cellular compartments) plays major roles in directing tyrosine nitration [170-175]. However, it is still unclear what specific elements, amino acid sequences, or protein conformations weight the most in influencing tyrosine susceptibility to nitration.

Importantly, tyrosine nitration has been reported to disrupt the regulation by phosphorylation of many central mediators of cell signaling and not surprisingly has been implicated in the pathogenesis of many diseases. Recent examples include the nitration of dynamin I Tyr374 that interferes with vesicular exocytosis [176], the nitration of insulin receptor substrate (IRS) that disrupts insulin signaling downstream of the insulin receptor [177], the activating nitration of p53 [178], and the inhibitory nitration of MAP-kinases and Akt that inhibits insulin-stimulated translocation of GLUT4 to the plasma membrane [179]. Therefore, examples are many in support of tyrosine nitration being an important post-translational modification of major relevance for cell signaling and physiology.

Protein Tyrosine Halogenation

Tyrosines can be modified by hypochlorite and hypobromite that result from the oxidation of chloride and bromide anions by peroxidases, in particular, myeloperoxidase (MPO) and eosinophil peroxidase (EPO) [180, 181]. MPO and EPO are expressed at high levels in specialized immune cells (e.g., neutrophils, eosinophils, and basophils). Therefore, tyrosine chlorination and bromination have been mostly appreciated in the context of granulocyte activation. Chlorine and bromine are relatively bulky electron withdrawing agents and produce tyrosine modifications more or less similar to nitration. Importantly, most of the HOCl and HOBr produced under biologically relevant conditions (<99%) is expected to be rapidly consumed by methionine and thiols making tyrosine halogenation an event of lesser importance except in very specific cases [182, 183]. Because the yield of tyrosine halogenation in vivo is low, it is predictable that under most relevant conditions minor impacts in cells signaling and function will result from these modifications although chloro- and bromo-tyrosines can serve as biomarkers of neutrophilic, eosinophilic activity, and oxidative stress [184].

Protein Tyrosine Hydroxylation

Tyrosine hydroxylation by ROS (hydroxyl radical) produces L-DOPA and m-hydroxy tyrosine as the terminal stable products [185]. L-DOPA is an intermediate in the synthesis of dopamine and epinephrine that is normally generated by tyrosine hydroxylase. In proteins, tyrosine hydroxylation produces protein bound L-DOPA or meta-hydroxy-tyrosine depending on whether hydroxylation occurs at the ortho or meta position relative to the phenolic hydroxyl group [185]. Hydroxylated tyrosines can redox cycle thereby propagating free radical reactions and amplifying oxidative stress [186-188]. The yield of protein hydroxylation nevertheless is generally low even under conditions of intense and persistent oxidative stress [185]. This is because, in vivo, hydroxyl radicals (the major source of uncatalyzed hydroxy-tyrosine formation) are produced only under fairly acidic conditions from peroxynitrite decomposition in the absence of thiols [158] or H2O2 reacting with redox-active metal ions (Fenton reaction [189]). Because of the ubiquity of CO2 that prevents peroxynitrite homolysis and the fact that redox active metal ions are not readily available, the production of hydroxyl radical is limited to localized events happening under very specific conditions. The lack of specificity in the reactions of hydroxyl radicals further limits tyrosine hydroxylation in vivo.

Tryptophan Oxidation

The oxidation of tryptophan (Trp) leads to the production of protein-bound kynurenine-derivatives, including N′-formyl-kynurenine. Recently, the development of a specific antibody to detect the formation and fate of the N′-formyl-kynurenine [41] made it possible to ascertain the importance of this modification of tryptophan in heme catalyzed photooxidation of α-crystallin, a process that may be of relevance in the onset and progression of cataract [42]. In addition, using immuno-detection of N′-formyl kynurenine, Ehrenshaft and colleagues showed the formation of oxidized proteins in mitochondria of UV-light exposed keratinocytes. Tryptophan oxidation has also been linked to the metal-catalyzed oxidation of LDL [190] and in the free-radical mediated crosslink of Cu,Zn-superoxide dismutase (SOD1) [191]. Taken together, recent advances indicate that the oxidation of tryptophan may be of importance in the mechanisms of pathogenesis initiated or promoted by oxidative stress resulting particularly from photochemical processes.

Concluding Remarks

Biological oxidations were initially investigated as detrimental consequences of exposure to ionizing radiation. Gradually, it became clear that oxidative modifications occur in normal physiology and regulate a variety of homeostatic processes necessary for the normal functioning of cells and tissues. As occurs with any process with such fundamental implications, the derailment of redox signaling has been linked to the onset and progression of most disease states. Unraveling the details of redox signaling in many different contexts is, thus, pivotal for the manipulation of specific pathways that in principle can be up or downregulated (pharmacologically and medicinally) to restitute health. For this, it is critical not only to continue to define sources and fates of ROS/RNS but also to focus on the determination of the identity and locality of their production in cells and tissues. This utter necessity is exemplified by exciting advances achieved by the design of “smart” antioxidants targeted to specific organelles [192-197] which showed in diverse models robust effects in comparison to the less effective generalized utilization of antioxidant therapies tried in the past. These exciting novel ways to manipulate redox signaling in cells converge with the many examples of very specific oxidations that change the course, attenuate, or potentiate signaling events. As the field conquers new grounds, it glimpses at yet more ambitious goals. The necessity to identify along with quantifying reactive species will eventually be satisfied by the re-evaluation, scrutiny, and creation of methodologies that will permit the real time in situ assessment of specific ROS/RNS generation in complex systems. Some of these are being implemented [60, 198]. Efforts in determining the limitations of available techniques to detect and quantify reactive species have contributed immensely to put many earlier findings in perspective and exemplified how critical scrutinizing the methodology is to successfully define the origins and effects of ROS/RNS in complex biological systems. It is critical then that methodologies that present limitations are replaced by new and improved techniques as promptly as they become available. As very briefly summarized in this review, much has been unraveled about the implications of biologic oxidation in health and disease. From these many exciting studies, it is possible to conclude that further advancing our comprehension of electron transfer reactions in biology will produce major impacts in preventive, regenerative, and curative medicine in the years to come.

Acknowledgements

The Bonini lab is funded by grants from the U.S. Department of Defense (W911NF-07-R-0003-04) and American Heart Association (13GRNT16400010). PCH is supported by NIH grant (5T32HL072742-09).

Ancillary