Shinya Toyokuni, MD, PhD, Department of Pathology and Biology of Diseases, Graduate School of Medicine, Kyoto University, Sakyo-ku, Kyoto 606-8501, Japan. E-mail: email@example.com
Aerobes, including humans, are consistently exposed to oxidative stress by consuming oxygen. The biological significance of oxidative stress via reactive oxygen and nitrogen species consists of two stages: reversible redox regulation and irreversible oxidative molecular damage, which are sometimes intermingled. During the past decade, many signaling cascades associated with oxidative stress have been discovered. An interaction between Keap1 and the Nrf2 transcription factor is among the most fundamental mechanisms of the defense system against oxidative or similar stress. Furthermore, it became apparent that reactive oxygen species are actively produced through enzymes such as xanthine oxidoreductase and nicotinamide adenine dinucleotide phosphate, reduced (NADPH) oxidases in non-phagocytic cells as well. The role of α-tocopherol solely as an anti-oxidant was also questioned. Now there is a long list of pathological states implicating oxidative stress. At the same time, genome projects on various species have been completed. These efforts convincingly led to a new era of oxidative stress investigation, contributing powerful strategies to select candidate genes or biomolecules. Herein are reviewed recent advances and novel concepts in this field, including oxygenomics. These fruitful results may lead to more accurate and useful pathological diagnosis and more efficient prophylaxis and therapeutic interventions on human diseases.
The longevity of the Japanese people has been continuously prolonged ever since World War II, and averaged 79 years in men and 86 years in women in the world health report from 2006 (http://www.who.int/whr/2006/en/index.html), and are world records. The top three causes of death in Japan were cancer, coronary heart disease and cerebral vascular disease in 2005.1 Of note is the fact that all these diseases are closely associated with oxidative stress. Under the current near-optimal environmental conditions, we may have to reconsider the ‘double-edged sword’ aspect of oxygen that may be regulating longevity.
Oxygen is indispensable for aerobes including humans. We utilize absorbed organic compounds via foods in combination with oxygen to produce energy. This process, called oxidative phosphorylation, uses the ability of O2 to be reduced to H2O via 4-electron reduction. These catalytic reactions are efficient but unfortunately not 100% efficient, because certain unexpected bypassed reactions produce reactive oxygen species in mitochondria. It is believed that approximately 2–3% of oxygen taken up will become a toxic intermediate.2 Furthermore, it has recently become clear that cells have mechanisms to positively produce reactive oxygen species either in the cytoplasm or on the plasma membrane via enzymatic reactions with a number of different mechanisms (Table 1). Here, we introduce two enzymes, xanthine oxidoreductase (XOR) and the nicotinamide adenine dinucleotide phosphate, reduced (NADPH) oxidase family. The former works in normal conditions as a xanthine reductase, reducing nicotinamide adenine dinucleotide+ (NAD+) to nicotinamde adenine dinucleotide, reduced (NADH). However, this enzyme, in case of either intramolecular disulfide formation3 or scission of the linker portion14 under oxidative stress, is converted to xanthine oxidase, and has been associated with the pathology of ischemia–reperfusion injury,15 producing either O2– or H2O2. Members of the NADPH oxidase (Nox) family, Nox2 and gp91phox, have long been recognized as the defense mechanisms against foreign materials and all kinds of infection in macrophages and B lymphocytes. However, this recognition is changing quite rapidly due to the presence of this enzyme in many other kinds of cells including endothelial cells.5 Another family protein, Nox1, is expressed abundantly in thecolonic epithelial cells, playing a role in immunity,16 but in vascular smooth muscle cells Nox1 is induced by angiotensin II. Indeed, in Nox1 knockout mice, hypertension is not induced by continuous administration of angiotensin II.4
Table 1. Enzymes producing reactive oxygen species
Aerobes are indeed cells that have succeeded in developing anti-oxidant defense mechanisms, and thus they are not stressed under ordinary conditions. However, there are many factors that greatly increase oxidative stress. Indeed, oxidative stress is associated with an array of pathological phenomena, including infection, inflammation, ultraviolet and gamma radiation, and overload of transition metals and certain chemical agents as well as ischemia–reperfusion injury.2Table 2 summarizes an updated list of pathological conditions associated with oxidative stress.
Table 2. Selected reports on pathological conditions associated with oxidative stress
As for the top cause of death in Japan, many epidemiological studies have demonstrated a close correlation between chronically oxidative conditions and carcinogenesis. For example, chronic tuberculous pleuritis causes a high incidence of malignant lymphoma;67 repeated exposure to asbestos fibers (crocidolite and amosite asbestos fibers are rich in iron68) is often associated with mesothelioma and lung carcinoma;69 chronic Helicobacter pylori infection is associated with a high incidence of gastric adenocarcinoma and lymphoma;42,70 the incidence of colorectal cancer is increased in inflammatory bowel diseases;71,72 a high risk for hepatocellular carcinoma is observed in patients with genetic hemochromatosis;73,74 severe burn by ultraviolet radiation is a risk factor for skin cancer;66,75 and gamma radiation causes a high incidence of leukemia.76 At least in the mentioned situations, and probably also in other more common types of carcinogenesis, oxidative stress appears to play a role.
CURRENT UNDERSTANDING OF OXIDATIVE STRESS
Oxidative stress depends on a balance between the load of reactive species and the anti-oxidant systems responsible for them. The latter contains (i) protective mechanisms for the initiation of the free radical reaction; (ii) what are called ‘anti-oxidants’ that efficiently retard the free radical reaction after its initiation; and last, (iii) repair mechanisms for modified or damaged biomolecules. There are two seamless stages in the biological significance of oxidative stress both historically and experimentally (Fig. 1).
The history of free radical chemistry dates back to 113 years ago. Fenton, a British chemist, reported in 1894 that ferrous sulfate and hydrogen peroxide cause the oxidation of tartaric acid, resulting in a beautiful violet color on the addition of caustic alkali.77 This was the basis for the discovery of the Fenton reaction, which produces hydroxyl radicals, presently known as the most reactive endogenous chemical species in living cells. In 1954 Harman proposed a hypothesis of ‘free radicals as a cause of aging’ for the first time.78 A decade was needed to integrate this idea into science, when McCord and Fridovich discovered superoxide dismutase, the first enzyme whose substrate is a free radical, in 1969.79 For the next two decades, studies based on the hypothesis that reactive species are damaging biomolecules were planned, and those efforts led to pioneering works on anti-oxidants, anti-oxidant enzymes, and protein oxidation,80,81 as well as DNA modification.82–84 Thus, the concept of ‘oxidative stress’ in the original sense was established.85
During the 1990s more scientists joined this research area using molecular biology techniques, and low levels of reactive oxygen and nitrogen species were established as signaling molecules in the cell.86,87 Multiple repair enzymes for oxidative DNA damage were also identified;88,89 nitric oxide (NO), a vasodilator, was integrated into free radical and oxidative stress research with reactions of NO and O2–90 and semiquantitative morphological analysis of oxidative damage was advanced.91 Indeed, reactive species were found to have two completely different characteristics that are seamless and sometimes intermingled, depending on the dose loaded and the situation each living cell or tissue has faced. When the reactions involved are reversible, it is described as redox regulation. In contrast, when the reactions are at the irreversible stage, they are designated oxidative molecular damage (Fig. 1). Aging may be a shift from redox regulation to oxidative damage.92
REDOX REGULATION AND SIGNAL CASCADES UNDER OXIDATIVE STRESS
It is not surprising that the antagonizing system for oxidative stress is reduction because oxidative stress oxidizes biomolecules. There are basically two major reductant systems, namely glutathione (GSH) and thioredoxin (TRX). A tripeptide, GSH was reported in the 1920s,93 whereas mammalian TRX was reported in 1988.94 Both of the systems constitute an array of associated enzymes and they work together at different levels. Although the intracellular concentration of TRX is 1/1000 that of GSH, TRX is essential for development because absence of this is lethal in TRX knockout mice.
Redox regulation is one of the key mechanisms for adaptation to a variety of stresses, including oxidative stress.95,96 Recently, it was reported from several independent laboratories including the authors' at Kyoto University that an antagonizing protein for TRX (thioredoxin-binding protein-2, TBP-2; also known as vitamin D3 upregulated protein-1, VDUP-1)97 is downregulated in cancers including human adult T-cell leukemia,98,99 human gastric cancer100,101 and iron chelate (ferric nitrilotriacetate)-induced renal cell carcinoma of rats.102 The mechanism of its inactivation was consistently methylation of the promoter region. Furthermore, TBP-2 is expressed at higher levels in non-metastatic melanomas than metastatic melanomas.103 Studies of a TBP-2-null mouse104 provided evidence that loss of TBP-2 results in enhanced sulfhydryl reduction and dysregulated carbohydrate and lipid metabolism, namely hyperinsulinemia, hypoglycemia, hypertriglyceridemia and increased levels of ketone bodies, at least in the liver and pancreatic β-cells.105 This was confirmed by producing TBP-2-deficient mice.106 The loss of TBP-2 appears advantageous to cancer cells because it ultimately results in facilitation of the glycolytic pathway via increasing the TRX activity. This makes sense in consideration of the fact that cancer cells are usually exposed to a lower oxygen environment.107 It is interesting to further note that the vitamin D3 receptor is one of the targets of p53-mediated transcriptional activation.108
One of the most striking findings during this decade is that many of the anti-oxidant enzymatic systems are under the control of a transcription factor, Nrf2.109,110 The cytoplasmic sensor for oxidative stress is now established as Keap1, a partner protein of Nrf2.111 In normal conditions, Nrf2 is bound to Keap1 in the cytosol and continuously degraded via the ubiquitin–proteasome system as soon as Nrf2 is translated. However, when oxidative stress is enhanced, Nrf2 is released from Keap1,112 translocates to the nuclear fraction, and binds to the anti-oxidant-responsive element, initiating the transcription of a variety of anti-oxidant enzymes including the GSH and TRX systems (Fig. 2). The target genes include (i) phase II detoxifying enzymes such as GSH-S-transferase and NAD(P)H:quinone oxidoreductase; (ii) GSH-associated enzymes such as γ-glutamylcysteine synthase and GSH peroxidase; and (iii) anti-oxidant enzymes such as heme oxygenase-1, peroxiredoxin, ferritin and TRX. Of note is the fact that this system can respond immediately to oxidative stress without producing new proteins. The cost of this system is the consumption of energy. Namely, the cells are consistently producing Nrf2 that is instantly decomposed via the ubiquitin–proteasome system.113 This appears to be a waste of energy, but this mechanism is essential for the fastest response against oxidative stress.
Here we have to point out the interaction between redox signals and classic phosphorylation signals. Regulation of phosphorylation of proteins with protein kinases and phosphatases is one of the most fundamental systems that living cells acquired during the evolution process. Some kinases are activated after exposure to oxidative stress;114 protein kinase C δ is activated via phosphorylation of Tyr512 by another kinase, c-Abl;115 the phosphatidylinositol-3 kinase-Akt(protein kinase B) pathway is also activated;116 apoptosis signal-regulating kinase-1 (ASK1), a mitogen-activated protein kinase kinase kinase, is also activated, further activating the c-jun N-terminal kinase and p38 pathways.117 Cells derived from ASK1 knockout mice are resistant to oxidative stress and septic shock by lipopolysaccharide.118 Interestingly, in the control condition, ASK is bound to TRX, producing an ASK1-signalsome where ASK1 is in the form of a homooligomer and resistant to phosphorylation. Oxidative stress releases ASK1 from the signalsome in parallel with the oxidation of TRX, leading to activation of the p38 pathway.119 It is now known that Toll-like receptor 4 is upstream of this pathway, with ASK1 pathway closely associated in natural immunity.120
Currently there has been great interest in the generation of O2– by the Nox family in association with redox signaling (Table 1). As has been demonstrated in the structural study of Nox, O2– is released to the outside of the cell or intracellular compartments. However, O2– cannot easily move through the membrane. It is assumed that O2– is converted to H2O2 through the dismutation reaction without the enzyme and taken up into the cytosolic compartment. One of the aquaporins (water channels) can take up H2O2, and this mechanism may also be utilized.121 Recently, it was reported that generation of reactive oxygen species through Nox1 is essential for cellular transformation by the ras oncogene.122 Furthermore, the involvement of Nox4 was reported in the signaling pathway of adipocytes after insulin stimulation.123
Even the role of vitamin E (α-tocopherol) as an anti-oxidant was questioned. It can at least suppress the activity of protein kinase C via activating protein phosphatase.124 There is a possibility that α-tocopheryl phosphate is a more efficient biomolecule.125 We recently observed that one of the anti-oxidant mechanisms of vitamin E is via inducing chaperone proteins associated with glycoproteins.126
A great advance was achieved with the understanding of the two gaseous mediators of oxidation, NO and carbon monoxide (CO). NO is enzymatically produced by nitric synthase from l-arginine, and there are basically two types of enzymes, constitutive and inducible. NO was identified as an endothelium-derived relaxing factor.127 There are three types of modifications by NO: S-nitrosylation (-S-NO) of cysteine residues in proteins, nitration (-NO2) of tyrosine residues in proteins and iron nitrosylation or chelation to heme. Activation of guanylate cyclase by NO is indeed achieved by its chelation to heme. These NO-associated modifications modulate protein interactions.128 For example, the interactions between HIF-1α (a transcription factor modulating responses for hypoxia) and p300129 or glyceraldehyde-3-phosphate dehydrogenase (a glycolytic enzyme) and Siah1130 are enhanced with these modifications, whereas those between N-ethylmaleimide-sensitive factor (NSF) and soluble NSF attachment protein receptor are compromised, relieving endothelial injury.131 Another important point is that in the presence of O2–, NO can produce a reactive species, peroxynitrite.90 This is the basis for the reaction of nitration in which 3-nitrotyrosine and 8-nitroguanine are produced. It was also hypothesized that nitrotyrosine-containing peptide with an affinity to SH2 domain of src kinase can activate src enzymatic activity by halting autoinhibition with intramolecular interaction to expose the catalytic domain.132
The majority of CO is produced via heme oxygenase in living cells, leading to the formation of CO, Fe(II) and biliverdin. This enzyme is abundantly present in the liver, brain, spleen, testis and placenta and is induced by oxidative stress. It is now known that CO has anti-inflammatory and cytoprotective activities,133 as well as modulatory actions on vascular tone.134 It is believed that low vascular tonus in the hepatic sinusoid is maintained by the presence of high concentrations of CO. Currently, even H2S has been recognized as a gaseous mediator.135
Oxygenomics, a new member of the omics family, is a synthetic word combining oxygen and genomics. We have recently proposed this research area136 in the hope that the biased distribution of oxidative DNA damage in the genome is associated with pathology of oxidative stress, especially in carcinogenesis, for the following reasons: the generation of excess reactive oxygen and nitrogen species causes DNA strand breaks, cross-links and modifications,82,84,91,137 leading to alterations in the genome information in spite of robust counteractions by repair enzymes and apoptotic pathways.138 These events may cause persistent activation of oncogenes or inactivation of tumor suppressor genes.139
Most scientists assume that free radical reactions present little specificity based on in vitro studies, in contrast to the extremely selective antigen–antibody interactions. Indeed, the second-order rate constant for the reaction of a hydroxyl radical with the DNA base, guanine is approximately 1.0 × 1010/s mol/L. Thus, one might consider that the genome is damaged at random and that there are no specific target genes or signaling pathways in oxidative stress-associated carcinogenesis. However, it is time to reconsider this assumption. We have been challenging this hypothesis recently, because ferric nitrilotriacetate (Fe-NTA)-induced renal cell carcinomas are homogeneous in histology.91,140 At the early stage of this rat carcinogenesis model, increased amounts of oxidatively modified DNA bases including 8-oxoguanine141 and a major lipid peroxidation product, 4-hydroxy-2-nonenal, and its modified proteins81,142–144 are observed.
In order to clarify whether there is any target tumor suppressor gene, we used a genetic strategy using microsatellite analysis in F1 hybrid rats145 and recently used comparative genomic hybridization (S. Akatsuka and S. Toyokuni, unpubl. data, 2007). This study found that the p15INK4B(p15) and p16INK4A(p16) tumor suppressor genes are two of the major target genes with a high incidence of homozygous deletion as well as methylation of the promoter region. This was the first report that showed the presence of a target gene in an oxidative stress-induced carcinogenesis model.146 The significance of this finding is enormous, especially because p16 is associated not only with the retinoblastoma protein pathway as an inhibitor of cyclin-dependent kinases 4 and 6, but also with the TP53 pathway via p19ARF and MDM2.147 p19ARF is an alternatively spliced transcript from the p16 tumor suppressor gene.148 Indeed, iron-mediated oxidative damage appears to attack one of the most critical loci of the genome. We then showed that allelic loss of p16 occurs as early as 1 week after the start of the animal experiment and is p16 gene-specific.149 Thus, we believe that these results suggest the presence of fragile sites in the genome. Elucidation of this mechanism and principle would change the chemopreventive strategy in particular cases of carcinogenesis.
Studying the localization of oxidative DNA damage in comparison with genome information and chromosomal structure is becoming increasingly important. There are numerous reports published on oxidative DNA damage in vitro using purified DNA or cultured cells. Based on these data, it has been claimed that certain specific sequences such as telomeres150,151 are especially vulnerable to oxidative damage. Currently, however, limited data are available on which part of the genome is susceptible to oxidative damage in vivo at the tissue or organ levels.
This is the post-genomic era, given the completion of genome projects of humans, mice, rats and other species (http://www.ncbi.nlm.nih.gov/Genomes/). The information in public databases is a great advantage for researchers of oxidative stress. We recently developed an immunoprecipitation-based method to make libraries of approximately 1 kb DNA fragments136 that contain one or more 8-oxoguanine152 or acrolein-modified adenine residues153 (Fig. 3). This is a versatile method, allowing the evaluation of relative content of selected oxidative DNA base modifications at any genomic locus when used in combination with polymerase chain reaction. Nuclear genomic DNA is integrated into the chromatin structure in association with histones. Some parts of the chromatin structure are open for transcription, whereas others are closed, constituting heterochromatin. Furthermore, genomic information, a blueprint for a cell, is not continuous information, but is divided into many pieces by the existence of chromosomes, although we cannot routinely observe chromosomal structure at interphase. Recently, the concept of chromosome territory has been established.154,155 This concept indicates that genome information corresponding to each chromosome is located at fixed areas, namely central nuclear or peripheral nuclear, in the nucleus even at the interphase. The localization appears to be different among different kinds of cells.156 This concept has been reviewed in a recent issue of Nature.157 It is possible that the genome areas susceptible to oxidative stress may differ depending on the kind of cells and the situation in which the cells are placed. Such a difference would help to explain the different signaling pathways each type of cancer has acquired, because not a few signaling pathways start with the recognition of DNA damage. We propose to call this novel research area ‘oxygenomics’. Oxygenomics is defined as the study of the localization of oxidative DNA damage in the genome in living cells. Using a DNA immunoprecipitation method, we found that the distribution of modified DNA bases is not random. Rather, the distribution of 8-oxoguanine and acrolein-modifed adenine was contrasting. Interestingly, aldehyde-modified DNA fragments were present in proximity to the nuclear membrane.136 This may be explained by the fact that acrolein, an aldehyde, comes from the cytoplasm or directly from the nuclear membrane.
PROTEIN MODIFICATION BY OXIDATIVE STRESS
The functions of proteins are finely regulated with a variety of modifications such as activation by proteolytic enzymes, modulation of activity and interaction by addition of sugars and phosphates, and localization to the membrane by palmitoylation or S-acylation.158 It is quite rare in eukaryotes that translated proteins function as they are, but rather they are subject to a variety of modifications after leaving ribosomes. Thus, post-translational protein modifications are essential for cellular activity.
In addition to these physiological modifications, proteins are subject to a variety of modifications by reactive species in pathological conditions; side chains of amino acids are directly oxidized (Fig. 4a) or modified by aldehydes that are secondary products of lipid peroxidation (Fig. 4b,c). These oxidative modifications may inactivate enzymatic functions and cause degeneration of structural proteins, but in contrast they may start activation of transcription factors or proteolytic systems. Thus, it is now recognized that comprehensive analyses of oxidatively modified proteins by the use of proteomics are important. This may be called oxyproteomics or oxidative stress proteomics. At least three different kinds of modifications are under investigation at present: oxidized cysteine, carbonyls and aldehydes.
Cysteine residues in proteins may work as a catalytic center and contribute to the formation of the stable 3-D structure of proteins by producing disulfide bonds with another cysteine residues within the same molecule. However, because sulfhydryl functions are quite susceptible to oxidative stress, they can be easily oxidized or subject to S-thiolation (Cys-S-S-X; Fig. 4a). This kind of modification occurs relatively early in oxidative stress, so they are early markers for protein modification (Fig. 1). There are two different methods for identification after 2-D gel analysis: one using the radioisotope 35S159 and the other using biotinylated cysteine.160 The latter has the merit of collecting cysteine-labeled proteins with avidin beads.
The most common method currently used for the detection of oxidatively modified proteins is the detection of carbonylated proteins.161 The principle of this method is the detection of dinitrophenylated proteins with monoclonal antibodies after the reaction of the carbonyl function (-C = O) with 2,4-dinitrophenylhydrazine. The merit of this method is its sensitivity, because carbonyl functions are contained not only in carbonylated amino acids but also in aldehydes and glycation products. In a recent work, target proteins were identified from cardiac and skeletal muscle of diabetic rats.162
Aldehydes are the major final products of lipid peroxidation, but are reactive enough to form covalent bonds with the side chains of certain amino acids (Fig. 4c). Recently it has become clear that these aldehydes can be second messengers for neighboring cells, generating hydrogen peroxide. Indeed, acrolein, 2-alkenal, 4-hydroxy-2-alkenal and ketoaldehydes (Fig. 4b) are among the aldehydes that have been associated with oxidative stress and cytotoxicity.81 We previously succeeded in producing monoclonal antibodies against 4-hydroxy-2-nonenal modified proteins.163 One of the major target proteins in the kidney after ferric nitrilotriacetate administration is β-actin.144 A recent report that modification of actin with electrophilic lipids alters its filament structure164 is consistent with our results. These kinds of studies are progressing quite rapidly, as shown in Table 3. These efforts would certainly lead to the discovery of new serum markers for oxidative stress-associated diseases.
Table 3. Target proteins of 4-hydroxy-2-nonenal modification
ApoE, apolipoprotein E; HNE, 4-hydroxy-2-nonenal; NA, not applied.
During the past decade not a few scientists came into this research area and it expanded a lot. However, this research area still has room for more expansion. The use of oxygen in life is a fundamental issue. Whether the surrounding environment on Earth gets better or worse, we cannot ignore the issue of oxygen. As the average lifespan is prolonged, the challenging problem for humans is the wearing-out of the blueprint of our system, the genome, and the other fundamental structures of the body. The fact that cancer and atherosclerotic diseases are the leading causes of death in most countries is the basis of this idea. At the same time, there are some emerging areas in science such as non-coding RNA.175,176 This is a rapidly expanding new field. So far, there are almost no reports on the association of oxidative stress and microRNAs. Also, epigenetics would be not less important for the pathological investigation of oxidative stress (Fig. 5).
Recently, a new technique was proposed to visualize free radicals by the use of an antibody against the 5,5-dimethyl-1-pyrroline N-oxide (DMPO)-nitorone spin trap form of molecules.177,178 This is an interesting idea because DMPO has been established as a probe for electron spin resonance. Most researchers call this research area oxidative stress research and we also used this term in the present review. However, we also have to return to the original idea of oxidative stress starting from the load of free radicals, especially from the point of spatial restriction of each reactive species. We hope that the present review article stimulates pathologists to work in the area of oxidative stress. Last, these fruitful results should be utilized to develop more accurate pathological diagnosis and more efficient prophylaxis and therapeutic intervention in the near future.
This work was supported in part by a MEXT grant (Special Coordination Funds for Promoting Science and Technology), a Grant-in-Aid from the Ministry of Education, Culture, Sports, Science and Technology of Japan, a Grant-in-Aid for Cancer Research from the Ministry of Health, Labour and Welfare of Japan, and a grant of Long-range Research Initiative (LRI) by Japan Chemical Industry Association (JCIA).