Oxidative stress in clinical situations - fact or fiction?

Abstract

The free radical story The appearance of oxygen into Earth's atmosphere 3 million years ago led to the death of many organisms which had been living up to that time in anaerobic conditions. However, this opportunity for oxidative metabolism provided an enormous metabolic advantage to aerobic organisms because the combustion of one molecule of glucose yielded 36 molecules of adenosine triphosphate (ATP), highly energetic substances. As the price for this metabolic advantage, surviving organisms were, however, obliged to develop antioxidant systems of defence to resist (or to control) the toxicity of oxygen which results from its ability to generate reactive and highly oxidant intermediates, notably free radicals. Oxygen was discovered in 1775 by Joseph Priestley who even then suggested the possibility that oxygen could be a toxic gas. In the last century, the 'Oxygen Paradox' was probably described for the first time by Louis Pasteur when he showed experimentally that anaerobic organisms died rapidly after exposure to air containing the 20% oxygen indispensable for the life of aerobic organisms. In 1900, Gomberg demonstrated that organic free radicals, i.e. a form capable of independent existence possessing one or more unpaired electrons in its chemical structure, could exist in living systems [1]. In 1934, Haber and Weiss [2] proposed that hydrogen peroxide could yield hydroxyl radicals (OH.) in the presence of iron salt acting as a catalyst agent as previously suggested by Fenton [3] in 1894. As early as 1946, Michaelis discussed the possibility that the formation of hydroxyl radical and hydrogen peroxide may occur during normal oxidative metabolism of molecular oxygen, through a univalent transfer of electron to the molecule [4]. In 1954, Rebeca Gerschman [5] presented supporting evidence that oxygen poisoning and X-ray irradiation with an animal model have one common mechanism of action, the formation of oxidizing free radicals. One year later, Denis Harman (nominated for the 1995 Nobel Prize for Medicine) expressed the theory that the aging process was associated with an abnormal metabolism of oxygen, leading to toxic free radical reaction [6]. In 1960, Laborit et al.[7] attracted attention to the possible toxicity of pressured oxygen in divers and suggested the use of some antioxidants for protection. In 1966, Slater evidenced that free radical reaction resulting from the hepatotoxicity of carbon tetra-chloride, was strongly associated with the generation of tissue injury [8]. On the recommendation of Horwitt [9], the Food and Nutrition Board in the USA officially recognized (1968) vitamin E, discovered by Evans and Bishop in 1922 [10], to be an essential vitamin because of its potential antioxidant properties. One year later, the discovery by McCord and Fridovich [11] of superoxide dismutase (SOD), enzyme present in erythrocytes and capable of destroying superoxide anion radical, started an explosive rush to explore the role of oxygenated free radicals in the development of human diseases. For further information, we advise the reader to consult reviews [12,13,14] or specialized books [15,16,17,18,19]. What is oxidative stress? When not correctly metabolized, oxygen gives rise to the formation of reactive oxygen species (ROS) including free radical species (superoxide anion, hydroxyl radical, peroxyl and alkoxyl radicals) or not (hydrogen peroxide, singlet oxygen, hypohalous acids). Physiological metals such as iron or copper acting as catalyst agents are strongly associated with the free radical chemistry [20,21]. From in vitro studies, it is well demonstrated that ROS are cytotoxic agents as a result of their ability to inactivate proteins, to oxidize lipoproteins, to promote DNA strand scission, to degrade carbohydrates, to induce haemolysis and overall to initiate lipid peroxidation processes. [for review, see 18]. Recently, many investigations have been devoted to the toxicity of reactive nitrogen metabolites with the celebrated free radical nitric oxide NO. [for review, see 17] but also of sulphur-centred radicals. Incubation of normal cardiac myocytes with a lipid peroxide results in abnormal contraction and membrane damage seen by 'bleb' formation on the cell surface [22]. Infusion of a free radical-generating system as iron-ADP complexes into the arterial supply of in situ canine diaphragmatic strip preparation results in profound alteration of the contractile function of the diaphgram [23]. In vivo, exposure of animals to 100% oxygen for 24 to 72 h, a well-known condition for increasing ROS production [24], causes measurable tissue injury and even death in rats [5,25]. Intranigral iron infusion into rat leading to lipid peroxidation formation in the brain was associated with progressive decrease in striatal dopamine, progressive atrophy of the substantia nigra and an increase in apomorphine-induced rational behaviour [26]. Under healthy conditions, the human body is continuously submitted to ROS but a large battery of antioxidants perfectly regulates the likely harm [27]. Indeed, ROS may play an important physiological role as illustrated with the killing of bacteria by granulocytes and macrophages [28], the regulation of blood arterial pressure by NO.; [29], or as recently linked to fertilization [30]. However, if free radical activity suddenly increases as the consequence of either a primary (e.g. excess radiation exposure) or a secondary (e.g. tissue damage by trauma) event, the antioxidant defence will be weakened, and in case of too prolonged ROS production, rapidly overwhelmed. As defined by Sies [31], an 'oxidative stress' occurs because there is a profound disturbance in the proxidant-antioxidant balance in favour of the former leading to potential for damage. Actually, ROS have been implicated in over 100 pathophysiological conditions, from cancer, arthritis, organ transplantation to AIDS, ageing and neurodegenerative diseases. To quote Halliwell [32], free radicals and antioxidants have become such 'hot news' that comments such as 'free radicals are implicated in Parkinson's disease'; 'taking vitamin E capsules will prevent heart attacks'; 'our anti-ageing cream contains free radical scavengers' are regularly found in the expert and lay press. However, despite enormous literature in this exciting field, most clinicians remain suspicious regarding the importance of the oxidative stress concept. This situation essentially arises from the fact that to measure free radicals reactions in vivo is a difficult challenge [33]. Clinicians have at their service automatic devices, which detect within a few minutes largely stable metabolites and enzymes in blood or urine samples, to determine functional disease in the body organs. This is not true for free radicals since the presence of the odd electron in their structure renders them extremely reactive with surrounding biological material and consequently not amenable to direct assay. In most if not all situations, the biochemist is consequently forced to determine much later after the clinical event secondary products of free radical reactions in peripheral blood, which may not necessarily identify the original locus of production. In this paper, a list of some analytical approaches available for free radicals species detection in vivo and an examination of their possible clinical utility is presented. How to monitor in vivo oxidative stress? In Fig. 1 the different methodologies are described which can be used to identify oxidative stress in human disease. They can be separated into nine categories: detection of oxidized biological compounds ('fingerprint assays'), salicylate hydroxylation, measurement of endogenous antioxidants, determination of total antioxidant capacity of blood, evidence of 'free iron', neutrophil activation, more sophisticated technologies such as bioluminescence and overall the electron spin resonance (esr) spectroscopy associated with the spin trapping technique, and finally anti-oxidant therapy.Fig. 1: Methodologies proposed for monitoring oxidative stress for clinical investigations.Detection of oxidized biological material Lipid damage. Generally present in high quantity in tissue, unsaturated fatty acids, phospholipids and cholesterol esters are the primary targets for free radicals resulting in the formation of lipoperoxides (LOOH) that can be detected in plasma by gas chromatography-mass spectrometry (GC/MS) [34] or by high performance liquid chromatography (HPLC) separation coupled with luminescence detection [35], both time-consuming techniques, or by using a commercially available assay kit (Kamiya Biomedical Co., Thousand Oakes, CA, USA) recently made available. Using one of these techniques, some investigators have reported the presence of lipid hydroperoxides in children with Kawasaki disease [36], in patients undergoing angioplasty [37] or developing Adult Respiratory Distress Syndrome (ARDS) [38] and in ischaemic hepatitis during circulatory shock [39]. However, great caution must be taken with LOOH determination since these are submitted to fast degradation both in vivo and in isolated samples generating byproducts such as aldehydes or alcanes. For many years, the most popular and easiest method but also the most controversial technique to reveal in vivo lipid peroxidation has been the spectro-photometric determination at 532 nm of malone-dialdehyde (MDA) by the thiobarbituric acid (TBA) assay [40]. Lack of specificity of the test (it is more appropriate to speak of TBA-reactive substances TBARS), the fact that MDA represents only a minor product of lipid peroxidation (less than 1%), in-terference with haemoglobin or biliverdin present in the sample, heating condition during the assay, presence of iron in reagents used for analysis, rapid metabolization of MDA are all conditions that render this technique totally artifactual so that most investigators recognized that 'it is scientifically unsound to equate increased plasma or serum levels of TBARS alone with the occurrence of a free radical disease [31,41]. Recently, attention has been focused on direct detection of MDA by HPLC method [42] which requires, however, meticulous care. A more specific method is the measurement of 4-hydroxynonenal (HNE), as the end product of lipid peroxidation, in plasma or red blood cells by HPLC equipped with a UV or fluorescence detector, a method proposed by Esterbauer et al.[43]. This methodology is relatively time consuming but affords trustworthy results. Several studies have reported increased levels of HNE in the development of human ischaemic hepatitis [39], and recently in patients undergoing liver transplantation [44]. Measurement of hydrocarbon gases, in expired breath, such as ethane or pentane (respectively products of peroxidation of n-3 and n-6 fatty acids) has been proposed as a specific, noninvasive and sensitive index of lipid peroxidation using many animal models of oxidative stress [45], but only in a few cases on man. Metabolism of pentane by the liver, the relative complexity of the equipment and all the care needed for a good sampling collection and for the detection of hydrocarbon gases can lead to pitfalls [46] in the interpretation of data (e.g. because of contamination of exogenous sources) this could explain why the method has not been extended to routine clinical studies. Using this technique, lipid peroxidation has been suggested to occur in volunteers exposed to oxygen or to physical exercise [47], in patients with myocardial infarction [48], alcoholic liver disease [49], rheumatoid arthritis [50], undergoing liver transplantation [51] and in one patient developing ARDS [52]. Protein and lipoprotein damage. Proteins are also sensitive to oxidative damage which leads to alterations in their primary, secondary and tertiary conformation which is reflected by changes in electrophoretic migration or in the fluorometric spectrum of the protein, by oxidation of thiol groups, and by loss of function. Two typical examples are the modified immunoglobulin G (IGg) and inactive alpha-1-proteinase inhibitor found in the synovial fluid from patients with rheumatoid arthritis [53,54]. Reaction of ROS with amino groups such as lysyl residues leads to the formation of a carbonyl group measured by the reaction with dinitro-phenylhydrazone (DNPH) as assessed by fluorescence spectroscopy or by the HPLC procedure [55]. Mono-and polyclonal antibodies against carbonyl-modified proteins are actually available [55]. In clinical situations suspected to be associated with heightened free radical activity, increased levels of DNPH derivatives have been found in the synovial fluid of patients with rheumatoid arthritis [55] and in plasma of patients with ARDS [56]. Protein thiol changes have also been reported in patients with ARDS [57]. The promotion of oxidative damage also leads to changes in lipoproteins and more particularly lowdensity lipoprotein (LDL) as evidenced in vitro and in vivo by the decrease in LDL polyunsaturated fatty acids and antioxidants, increased concentration of MDA and HNE, fragmentation of the apolipoprotein B to smaller peptides, increase in the electrophoretic mobility of oxidized LDL or increased uptake by the macrophage scavenger receptor [58]. Demonstrating that LDL oxidation can occur is an important in vivo phenomenon since it has been strongly related to the development of atherosclerotic occlusions. Immunocytochemical analysis of antibodies against oxidatively modified LDL is available for in vivo studies, but not commercially [59]. Recently, Salonen et al. reported the occurrence of autoantibodies to MDA-LDL in atherosclerotic Finnish men, thereby supporting a role for ROS in this disease [60]. DNA damage. Interaction of free radicals, and more especially hydroxyl radical, with bases of DNA gives rise to some derivatives (8-hydroxy-guanine, thymine glycol) that can be detected in urine after excision and restitution of the oxidized DNA [61]. However, their analysis is, delicate and requires pre-purification on immunoaffinity columns and the use of HPLC with electrochemical detection or GC/MS. With this technique, Bergtold and associates showed that species (rat, mouse) with higher metabolic rates and consequently exposed to a higher oxidative stress, exhibited in urine a 20 fold-increase in oxidative DNA damage, when compared to human urine [62]. In autoimmune diseases including rheumatoid arthritis, 8OHdG concentration was increased in circulating mononuclear cells isolated from patients [63]. Although considered as a good index of oxidative stress, determination of DNA damage is very time consuming and requires better analytical standardization so this technique is not easy to use in clinical situations. Salicylate hydroxylation This method is based on the ability of OH.; to attack the benzene rings in aromatic molecules and to produce hydroxylated products that can be quantified by the HPLC procedure. For in vivo studies, some important conditions are required: the hydroxylated product must not be metabolized and must be different from possible enzyme-produced hydroxylated metabolites. In vitro and in vivo studies in animals have shown that salicylic acid in its acetylated form (aspirin) was a suitable agent which resulted in the formation of 2, 3 dihydroxybenzoic acid (2, 3 dHB) as an oxidative byproduct. After oral administration of aspirin, elevated levels of 2, 3 dHB were found in the plasma of patients undergoing chronic oxidative stress, rheumatoid arthritis [64] and diabetes melittus [65], suggesting the involvement of OH. radicals in such pathogenesis. Measurement of endogenous antioxidants It is well-established that antioxidants are consumed in the in vitro model during oxidative stress. An example is given of the red blood cells exposed to AAPH (2,2′-azo-bis(2-amidinopropane hydrochloride) as an in vitro free radical generating system which promotes extensive haemolysis associated with a decrease in vitamin E (α-tocopherol) (Fig. 2). Many studies on animals have described the decrease in endogenous antioxidants in situations of oxidative stress, more particularly ischaemia-reperfusion [66,67].Fig. 2: Relation between decrease in red blood cell (RBC) levels of vitamin E and time course of haemolysis when cells are exposed to 100 mM water soluble azo-compound AAPH (2,2′-azo-bis(2-amidinopropane hydrochloride), as free radical-generating system. Haemolysis generally starts when vitamin E has fallen to ± 10% of initial values.Human plasma does not contain enzymatic anti-oxidants (other than superoxide dismutase with 5-20 U mL−1) but is rich in small molecular antioxidants being divided into water soluble (uric acid, ascorbic acid, bilirubin, glutathione) and lipid-soluble (α-tocopherol, ubiquinol-10, α and β-carotene). In contrast with plasma, red blood cells are very rich in enzymatic antioxidants (superoxide dismutase, glutathion peroxidase, catalase). In the human, significant depletion of such antioxidants have been reported in several pathophysiological conditions: rheumatoid arthritis [68], HIV infections [69], acute pancreatitis [70], liver and renal transplantation [41,71], ARDS [72], cardiopulmonary bypass procedure (CPB) for cardiac surgery [73,74], sepsis [75] and ischaemic hepatitis [39]. Our unpublished data [76] has revealed, in patients with severe head trauma, a loss of plasma vitamin E after their admission to hospital and this has recently been confirmed by Braughler and Hall [77]. From a general point of view, antioxidant determination appears to be indicative of the propensity of the individual for oxidative stress. Determination of total antioxidant capacity of blood This methodology is designed to determine the global ability of plasma to inhibit in vitro the free radical reaction by its protective defences. If its protective defences have been consumed during in vivo stress, it could be expected that the plasma will have lost a great part of its inhibiting capacity. The total radical-trapping capacity of plasma can be evaluated according to the TRAP (Total (peroxyl) Radical-trapping Antioxidant Parameter) assay initially developed by Wayner et al.[78]. Water-or lipid-soluble azo-compounds such 2,2′-azo-bis(2-amidino-propane hydrochloride) (AAPH) or 2,2′-azo-bis(2,4 dilethyl) valeronitrile (AMVV) are generally used as an in vitro peroxyl radical generator and the oxidative damage can be followed by measuring oxygen consumption, chemiluminescence or loss of fluorescence of the protein R-phycoerythrin (ORAC assay) [79] in plasma samples, or by monitoring haemolysis of red blood cells [80]. Using this technique, Mulholland and Strain [81] have observed that TRAP values were significantly lower in the plasma of patients after acute myocardial infarction (AMI) when compared with sexand age-matched controls, indicating a decrease in the plasma antioxidants in the AMI group. Erythrocytes of newborn babies have been shown to be more sensitive to oxidation by AAPH than those in normal babies [82]. Actually, a commercially available kit (Randox Total Antioxidant Status Kit, Antrim, UK) is based on interference of plasma by the radical cation ABTS.;+ generated from the incubation of ABTS and hydrogen peroxide in the presence of a peroxidase. Using this methodology, Miller et al.[83] showed that pre-term babies had a significantly depressed plasma antioxidant activity at birth when compared to that of term babies. Release of pentane, as marker for lipid peroxidation, has also been used to estimate the global antioxidant capacity of red blood cells or plasma exposed to in vitro free radical generating systems [84,85]. In patients undergoing CPB, Hartstein et al.[86] showed that the susceptibility of plasma to lipid peroxidation (induced by γ-irradiation of samples) dramatically increased after the start of CPB, indicating a consumption in antioxidants demonstrated by concomitant falls in vitamin E in the same plasma samples. Recently, Toivonen et al.[87] have confirmed these preliminary observations. TRAP assay is useful for a rapid screening of samples because techniques which use in vitro generation are easy to standardize. Nevertheless, TRAP always requires specific assays to determine which individual antioxidant has been affected. Evidence of free iron Transition metals, most particularly iron, play a key role in initiating free radical chemistry through the Fenton reaction. In the human, the amount of free iron in biological samples is negligible as almost all iron is physiologically stored in ferritin, lactoferrin and transferrin. Under the action of superoxide anion radical, it has been shown that iron could be released as free or in low molecular weight forms from stock proteins and then be able to participate in a free radical reaction [88]. Haemoglobin in the presence of hydrogen peroxide can also release its iron which is capable of initiating lipid peroxidation [89]. Therefore, detection of free iron can be considered to be a good index of in vivo oxidative stress. Several years ago, the bleomycin assay for chelatable redox active iron was introduced by Gutteridge et al.[90] to detect this in biological samples, as reported in synovial fluid of patients with rheumatoid arthritis [91] and recently in the plasma of patients undergoing the cardiopulmonary bypass procedure for routine aortic valve replacement [92,93]. In this last study, the authors also investigated the iron-binding ability of plasma transferrin which in normal subjects is only one third loaded with iron. Two assays were used to measure this iron-binding antioxidant activity of transferrin employing an organic oxygen radical and an oxo-iron system as the damaging or provocative system [94]. In CPB patients, this study showed that in 13% patients transferrin became saturated with iron by the end of bypass, therefore losing its antioxidant properties. Neutrophil activation When stimulated, the polymorphonuclear cells (PMNs) are an important source of activated oxygen species in biological systems. Their participation in the inflammatory phenomenon provides a link between inflammation and free radical-induced lesions. In addition to ROS, activated PMNs also release into the extracellular medium enzymes such as elastase and myeloperoxidase. Their plasma concentration detected by radioimmunoassay or Elisa tests may therefore be taken as specific markers of PMNs activation in pathological states and indirectly as a proof of increased ROS production. When compared with control groups, abnormal concentrations of such enzymes have been found in patients with septic shock [95] or in patients undergoing CPB [96], dialysis [97] or kidney transplantation [71]. Bioluminescence Production of ROS is related to the emission of light in the wavelengths located in the near-infrared and infrared (spontaneous chemiluminescence). Among ROS is singlet oxygen which it has been suggested contributes to the induction of lipid peroxidation and DNA strand scission. Relaxation of this electronically activated singlet oxygen to the ground state produces photon emission that can be detected at 1268 nm [98]. Detection of ultralow light levels (emitted photons being captured by tissue) requires a chemi-luminescence spectrometer equipped with a liquid-nitrogen-cooled germanium diode that provides high sensitivity to near infrared radiation from 1000 to 1700 nm and a singlet oxygen calibration standard obtained by the reaction of hydrogen peroxide with hypochlorous acid. Direct luminescence of centrifuged fresh human urine has been shown to be related to the amount of peroxides in urine [99]. In patients suffering from Duchenne Muscular Dystrophy (DMD), enhanced urinary spontaneous luminescence is associated with increased level of creatinine, a marker molecule for the disease [100]. In a recent pilot study on rats, Dirnagl et al.[101] have shown that in vivo spontaneous or lucigenin-enhanced chemi-luminescence were significantly increased in global forebrain ischaemia followed by reperfusion and in meningitis induced by pneumococcal suspension, when compared with control animals. For its application to in situ clinical studies, this technology is, however, limited by the size of the device but also by the absolute necessity to work in absolute darkness in order to avoid light interference. Electron spin resonance (esr) spectroscopy The chemiluminescence, technologies described above for measuring in vivo free radical production only give indirect results and, therefore, do not afford conclusive data about the exact role played by ROS in inducing tissue injury. Electron spin resonance (esr) spectroscopy is currently considered to be the technique of choice for detecting free radicals because it allows free radical detection by measuring the absorption of energy as a result of interaction by the unpaired electron present in the free radical with an applied external magnetic field produced by two magnets. As a result of the nature of the free radical, this interaction results in typical esr spectra that can be visualized on a computer screen, analysed, and even quantified since the height of the signal is directly proportional to the concentration of free radical present in the sample being analysed. Direct esr analysis is, however, of very limited application for in vivo use since esr methodology remains relatively insensitive and requires steady-states concentrations of free radicals in the micromolar range. In fact, only two radicals can be detected directly by esr in biological samples as complex as blood, plasma or tissue: ascorbyl radical and nitric oxide (NO.) radical. Ascorbyl radical derived from the oneelectron reduction of ascorbate has a prolonged life duration because of the delocalization of the unpaired electron on the whole molecule. Its detection by esr at room temperature in plasma has been proposed as a noninvasive marker for oxidative stress [102]. Examples have been given of clinical studies with patients undergoing aortic cross-clamping ischaemia [103]. The esr spectrum of NO., recognized as an endothelial derived release factor (EDRF), can also be directly identified in whole blood sample frozen at 77°K since its binds firmly to haemoglobin to form a stable species HbNO.[104]. In the animal model, increased formation of NO. is thought to be implicated in the development of ischaemic brain injury; this was demonstrated recently using esr by Tominaga et al.[105]. Following a 5-min period of leg ischaemia in human studies, Wennmalm reported increased esr signals of HbNO.; in plasma samples collected within the 5-10 min of reperfusion [106]. In order to obtain steady-state concentration of free radicals in the detectable range, the use of spin trapping agents (nitrones, nitrosones) is required for analysis of samples in vivo. These agents have indeed the ability to react with transient free radicals to form much more stable radicals (spin adducts) that are longer-lived than the original species. Numerous spin trapping agents are available: 5, 5-dimethyl-1-pyr-roline-1-oxide (DMPO), α-phenyl-N-tert-butylnitrone (PBN), 2-methyl-2-nitrosopropane (MNP), pyrridyl-N-oxide-tert-butylnitrone (POBN), 2,2,6,6-tetramethylpiperidine-1-oxyl (TEMPO). Spin trapping agents have some specific affinity for free radicals: DMPO, as hydrophylic agent, is useful for trapping superoxide anion or hydroxyl radical while PBN, a lipophylic molecule, preferentially reacts with lipid-derived radicals. In animal studies, spin trap agents are injected before inducing oxidative stress such as ischaemia-reperfusion [107,108,109], septic shock [110], hyperoxia [111], carbon tetrachloride administration [112]. Unfortunately, these agents have some toxic properties and cannot therefore be administered safely to human patients. Actually, this problem can be avoided by using a noninvasive spin trapping technique in which blood samples are drawn and mixed with the spin trap as soon as possible (ex vivo reaction). In the field of ischaemia-reperfusion, a first application of this technique was given by Coghlan et al.[113] who reported the appearance of an esr signal in the blood of a road-traffic injured man undergoing delayed repair of a transected aorta. After heart reperfusion, the same authors also detected esr signals in PBN-treated blood taken from the coronary sinus of patients undergoing percutaneous transluminal coronary angioplasty [37]. In patients undergoing elective open heart injury, Tortolani et al.[114] demonstrated a biphasic profile of free radical production in the coronary sinus blood with an initial rise from 5 to 10 min after heart reperfusion followed by a second peak at 25 min. PBN-adduct can also be observed in the peripheral blood of such patients [115]. Using DMPO as another spin trap agent, Culcasi et al.[116] identified an increase in DMPO-OH. formation in the coronary sinus blood drawn from patients subjected to cardiopulmonary bypass during the first minutes following aortic declamping. Recently, we obtalned first evidence of a growing free radical production in patients subjected to renal transplantation [71]. In order to reduce the time for mixing, surgeons had at their immediate disposal syringes containing freshly prepared and sterilized PBN as the blood samples were directly collected from the renal vein. By using this method, we recorded that a rise in free radical production occurred within the first minutes after the graft reperfusion. Antioxidant therapy Despite the complexity of measuring in vivo ROS, there are increasing numbers of studies which indicate, more particularly since the most frequent use of esr spectroscopy, that their generation significantly increases in certain human pathological conditions. Although the toxicological role of ROS is firmly established by in vitro studies and, to a less extent by animal