Oxygen-derived species: their relation to human disease and environmental stress.

Free radicals and other reactive oxygen species (ROS) are constantly formed in the human body, often for useful metabolic purposes. Antioxidant defenses protect against them, but these defenses are not completely adequate, and systems that repair damage by ROS are also necessary. Mild oxidative stress often induces antioxidant defense enzymes, but severe stress can cause oxidative damage to lipids, proteins, and DNA within cells, leading to such events as DNA strand breakage and disruption of calcium ion metabolism. Oxidative stress can result from exposure to toxic agents, and by the process of tissue injury itself. Ozone, oxides of nitrogen, and cigarette smoke can cause oxidative damage; but the molecular targets that they damage may not be the same.


Introduction
There is considerable current interest in the role of free radicals, oxygen radicals, and oxidative stress as mediators of tissue injury in human disease and of the effects of air pollutants, such as ozone (03), nitrogen dioxide (NO2), and tobacco smoke. This article explains the meaning of these terms and summarizes current knowledge of the roles they play in these various situations.
What Is a Radical?
In the structure of atoms and molecules, electrons usually associate in pairs, each pair moving within a defined region of space (an atomic or molecular orbital). One electron in each pair has a spin quantum number of +1/2, the other -1/2. A free radical is any species capable of independent existence (hence the term "free") that contains one or more unpaired electrons, an unpaired electron being one that is alone in an orbital (1). The simplest free radical is a hydrogen atom, with one proton and a single electron. Table 1 gives some examples of other free radicals. Note technique of electron spin resonance is often used to measure free radicals; it records the energy changes that occur as unpaired electrons align in response to a magnetic field. A superscript dot ( ) is used to denote free radical species.

Oxygen-A Free Radical and Environmental Toxin
When living organisms first appeared on the earth, they did so under an atmosphere containing very little 02, i.e., they were essentially anaerobes. Anaerobic microorganisms still survive to this day, but their growth is inhibited and they can often be killed by exposure to 21% 02, the current atmospheric level. As the 02 content of the atmosphere rose (due to the evolution of organisms with photosynthetic watersplitting capacity) many primitive organisms probably died. Present-day anaerobes are presumably the descendants of those primitive organisms that followed the evolutionary path of "adapting" to rising atmospheric 02 levels by restricting themselves to environments that 02 did not penetrate. However, other organisms began the evolutionary process of evolving antioxidant defense systems to protect against 02 toxicity. In retrospect, this was a fruitful path to follow since organisms that tolerated the presence of 02 could also evolve to use it for metabolic transformations (involving enzymes such as oxidases and oxygenases) and for efficient energy production (by using electron transport chains with 02 as the terminal electron acceptor, such as the mitochondrial oxidative phosphorylation system). Hence, 02 was probably the first environ-mental air pollutant to appear in large quantities on the planet.
However, even present-day aerobes suffer oxidative damage if they are exposed to 02 at concentrations greater than 21% (2).
Oxygen toxicity has been demonstrated in plants, animals and microorganisms. For example, exposure of adult humans to pure 02 at 1 atm pressure for as little as 6 hr causes chest soreness, cough, and sore throat in some subjects; and longer periods of exposure lead to lung damage. The incidence of ocular damage in babies known as retrolental fibroplasia ("formation of fibrous tissue behind the lens") increased abruptly in the early 1940s among babies born prematurely and led to many cases of blindness. Not until 1954 was it realized that retrolental fibroplasia is associated with the use of high 02 concentrations in incubators for premature babies. More careful control of 02 concentrations (continuous transcutaneous 02 monitoring, with supplementary 02 given only where necessary) and administration of x-tocopherol have decreased its incidence, but the problem has not disappeared, since many premature infants need continuous high 02 to survive at all (3).
The damaging effects of elevated 02 on aerobes vary considerably with the organism studied, age, physiologic state, and diet; and different tissues are affected in different ways. Thus, cold-blooded animals such as turtles and crocodiles are relatively resistant to 02 at low environmental temperatures, but become more sensitive at higher temperatures. Neonatal rats resist lung damage in an atmosphere of 100% 02 far more effectively than do adult rats (2). Environmental Health Perspectives The earliest suggestion made to explain 02 toxicity was that 02 is a direct inhibitor of enzymes, thereby interfering with metabolism. However, very few targets of direct damage by 02 have been identified in aerobes. In 1954, Gerschman et al. (4) proposed that the damaging effects of 02 could be attributed to the formation of oxygen radicals. This hypothesis was popularized and converted into the superoxide theory of 02 toxicity following the discovery of superoxide dismutase (SOD) enzymes by McCord and Fridovich (5). In its simplest form, this theory states that 02 toxicity is due to excess formation of superoxide radical (O2-)' the one electron reduction product of 02, and that the SOD enzymes are important antioxidant defenses because they remove p--. Ironically, with all the fuss made about oxygen radicals, it must be realized that the diatomic oxygen molecule is itself a free radical, containing two unpaired electrons (1). Fortunately, the electronic arrangement in 02 renders this molecule unreactive despite its free radical nature (1).

Reactive Oxygen Species in Vivo
Reactive oxygen species (ROS) is a collective term used by biologists to include not only oxygen radicals (O-p and hydroxyl radical, OH) but also some derivatives of 02 that do not contain unpaired electrons, such as hydrogen peroxide (H202), singlet 02(1Ag), and hypochlorous acid (HOCI)*.
Reactive is of course a relative term: Op is more reactive than 02, but neither Op nor H202 in aqueous solution is anywhere near as reactive as OH(1).
All organisms suffer some exposure to OH-, because it is generated in vivo by homolytic fission of 0-H bonds in water, driven by our continuous exposure to background ionizing radiation (6). Hydroxyl radical is so reactive with all biological molecules that it is impossible to evolve a specific scavenger of it-almost everything in living organisms reacts with OH with second-order rate constants of 109 to 1010 M -1sec l (essentially, if OH-contacts the compound, reaction occurs). Damage caused by OH-, once this radical has been formed, is probably unavoidable and is dealt with by repair processes ( Table 2).
It is now well established (1,5,(18)(19)(20) that 0p and H202 are produced in aerobes, although the precise amounts generated and the steady-state concentrations achieved are still uncertain. Generation of these species occurs by two types of processes described below. "Accidental" Generation. This encompasses such mechanisms as "leakage" of electrons onto 02 from mitochondrial electron transport chains, microsomal cytochromes P450 and their electron donating enzymes, and other systems (1,5,20). It also includes so-called autoxidation reactions in which compounds such as catecholamines, ascorbic acid, and reduced flavins are alleged to react directly with°2 to form O-- (5). In fact, such autoxidations are usually catalyzed by transition metal ions (1).
Deliberate Synthesis. The classic example of deliberate metabolic generation of ROS for useful purposes is the production of O2, HOCI, and H202 by activated phagocytes (21). Hydrogen peroxide is additionally generated in vivo by several oxidase enzymes, such as glycolate oxidase, xanthine oxidase, and D-amino acid oxidase (18,22). Evidence is accumulating that 0 -is also produced by several cell types other than phagocytes, including lymphocytes (23), fibroblasts (24,25), and vascular endothelial cells (26)(27)(28). Such Op might be involved in intercellular signalling and could serve important biologic functions, although more information is needed. nize abnormalities in DNA and remove attacks guanine preferentially. H202 them by excision, resynthesis, and rejoin-and°2-do not attack DNA (9).

Proteins
Many ROS can oxidize -SH groups. Oxidized methionine residues may be (12,13) OH attacks many amino acid residues repaired by methionine sulfoxide reduc- ( 11). Proteins often bind transition tase. Other damaged proteins may be recmetal ions, making them a target of ognized and preferentially destroyed by attack by site-specific OH' generation cellular proteases. (1,11).
Environmental Health Perspectives *HOCI could equally well be regarded as a "reactive chlorine species." Generation of O-, HOCI, and H202 by phagocytes is known to play an important part in the killing of several bacterial and fungal strains (21). Some other metabolic roles for H202 have been proposed (29)(30)(31)(32)(33). For example, H202 is used by the enzyme thyroid peroxidase to help make thyroid hormones (30). H202 or products derived from it can displace the inhibitory subunit from the cytoplasmic gene transcription factor NF-KB. The active factor migrates to the nucleus and activates genes by binding to specific DNA sequences in enhancer and promoter elements. Thus, H202 can induce expression of genes controlled by NF-KB. This is of particular interest because NF-KB can induce the expression of genes of the provirus HIV-1, the major cause of acquired immunodeficiency syndrome (33). H202, a nonradical, resembles water in its molecular structure and is very diffusible within and between cells.
Much O-p generated in vivo probably undergoes a dismutation reaction to give H202, as represented by the overall equation 20i-+ 2H+ 3-H202 + 02 [1] Toxicity of Superoxide and Hydrogen Peroxide Experimental data show clearly that removal of O-p and H202 by antioxidant defense systems is essential for healthy aerobic life (1,5,18,34). Why is this? In organic media, 0°can be very reactive but in aqueous media it is not, mainly acting as a moderate reducing agent, e.g., the reduction of cytochrome c cyt c (Fe)3++ Op-+ cyt c (Fe2+) + 2 [2] However, O-p, can react with some targets. In particular, O-p reacts fast with nitric oxide in a radical addition reaction (35).

2-p+ NO -* ONOO0
peroxynitrite [3] NOis known to be produced in vivo by vascular endothelial cells, by some cells in the brain, and by phagocytes (8). NO performs useful physiologic functions, such as regulation of vascular smooth muscle tone (hence controlling blood pressure) and neurotransmitter action (8). Since NO-acts upon smooth muscle cells in blood vessel walls to produce relaxation, then O2-, by removing NO-(Equation 3), can act as a vasoconstrictor, and this might have deleterious effects in some clinical situations (36,37).
Considerable debate continues in the literature as to whether or not the interaction of O-and NO (Equation 3) is damaging to cells (38). Peroxynitrite might be directly toxic to cells (38,39). It might also decompose to form a range of toxic products, including some OH- (38,40).
ONOO-+ H+ -OH-, NOQ, NO+ [4] Note that Equation 4 also produces the toxic free radical gas nitrogen dioxide. However, the physiologic significance of these reactions is still uncertain, since some experiments suggest that NO' may protect against oxidative damage even when O-p is being generated (41,42).
Superoxide has also been shown to be capable of inactivating several bacterial enzymes, such as Escherichia coli dihydroxyacid dehydratase, aconitase, and 6phosphogluconate dehydratase (5,20,43). It appears to attack iron-sulfur clusters at the enzyme active sites. Whether such reactions of Opoccur in mammals is uncertain, although in isolated submitochondrial particles, O-p has been claimed to inactivate the NADH dehydrogenase complex of the mitochondrial electron transport chain (44). The protonated form of O2hydroperoxyl radical (HO2), is much more reactive than O°in vitro. For example, HOpcan initiate peroxidation of polyunsaturated fatty acids and decompose lipid hydroperoxides, which O2cannot (45,46). However, there is no direct evidence that HOpexerts damaging effects in vivo. The pKa of HOpis about 4.8, so only a small fraction of Ois protonated at physiologic pH values.
It is usually thought that most or all of the toxicity of Oand H202 involves their conversion into OH' (1,48). Several mechanisms have been proposed to explain this.
Iron and copper (and other transition metal ions) in chemical forms that can decompose H202 to OH-are in very short supply in vivo: organisms are very careful to ensure that as much iron and copper as possible are kept safely bound to transport or storage proteins. Indeed, the "sequestration" of metal ions into forms that will not catalyze free radical reactions is an important antioxidant defense mechanism (1,29,53). Sequestration of iron and copper ions deters the growth of many bacterial strains in human blood plasma (54): it also ensures that plasma will not convert O2and H202 into OH' (29,53). Prevention of OH' formation may allow small quantities of O-p and H202 released into the extracellular environment (e.g., from endothelial cells, lymphocytes, and phagocytes) to perform useful metabolic roles, such as intercellular signaling, rather than causing damage (29).
In any case, any transition metal ions that do become available to catalyze free radical reactions in vivo will not exist in the "free" state for very long. Thus, if iron ions are liberated, they must bind to a biological molecule or else eventually precipitate out of solution as ferric hydroxides, oxyhydroxides, and phosphates. If metal ions bound to a biological molecule react with O-p and H202 (Equations 5,6), OH' will be formed upon the molecule. This OHmediated damage is said to be "site specific" (55). Binding of metal ions to a target means that any OH' generated will tend to react with that target rather than with any added scavenger, and the OHwill be very difficult to intercept by OH' scavengers.
It follows that a major determinant of the nature of the damage done by excess generation of ROS in vivo is the availability and location of metal ion catalysts of OH-radical formation (1,29). If, for example, "catalytic" iron or copper ions are bound to DNA in one cell type and to membrane lipids in another, then excessive formation of H202 and O°-will, in the first case, damage the DNA and in the other could initiate lipid peroxidation. Evidence for OH' formation in the nucleus of cells treated with H202 has been Volume 102, Supplement 10, December 1994 obtained, presumably involving metal ions bound upon, or very close to DNA (56,57). E. coli mutants lacking SOD activity are hypersensitive to damage by H202 (34), and extra SOD can often protect cells against damage by H202, provided that it can enter the cell (58). These data are consistent with a role of O2in facilitating damage by H202, and Equations 5 and 6 provide an explanation. However, many scientists are reluctant to believe that O°serves only as a reducing agent for metal ions since, in general, mammalian tissues are fairly reducing environments. The arguments have been rehearsed in detail (1,49,59) but the point is not yet settled.

Antioxidant Defenses
Living organisms have evolved antioxidant defenses to remove excess O-and H2O Superoxide dismutases (SODs) remove Oby greatly accelerating its conversion to H202 (Equation 1). Human cells have a SOD enzyme containing manganese at its active site (Mn-SOD) in the mitochondria. A SOD with copper and zinc at the active site (Cu,Zn-SOD) is also present, but largely in the cytosol (5). Catalases in the peroxisomes convert H202 into water and 02 and help dispose of H202 generated by peroxisomal oxidase enzymes (18).
However, the most important H202removing enzymes in human cells are glutathione peroxidases (GSHPX), which require selenium (as an active site selenocysteine residue) for their action. GSHPX enzymes remove H202 by using it to oxidize reduced glutathione (GSH) to oxidized glutathione (GSSG). Glutathione reductase, a flavoprotein enzyme, regenerates GSH from GSSG, with NADPH as a source of reducing power (18). Another important antioxidant defense already referred to is the sequestration of transition metal ions into forms that will not catalyze free radical reactions (1,29,53). This is particularly important in the extracellular environment, where levels of SOD, GSH, GSHPX, and catalase are often very low (53). Antioxidant defense enzymes are essential for healthy aerobic life. For example, SOD-negative mutants of E. coli will not grow aerobically unless given a rich growth medium, due to impaired biosynthesis of certain amino acids. Even when so supplemented, SOD E. coli cells grow slowly, suffer membrane damage, are abnormally sensitive to damage by H202 (perhaps because of Equations 5 and 6) and show a high mutation rate (34).
However, antioxidant defenses exist as a balanced and coordinated system. Thus, although SOD is important, an excess of SOD in relation to peroxide-metabolizing enzymes can be deleterious (60)(61)(62)(63). This has been shown by transfecting cells with human cDNAs encoding SOD (60). Transgenic mice overexpressing human Cu,Zn-SOD are resistant to elevated 02 and to certain toxic agents (62,63) but they show certain neuromuscular abnormalities resembling those found in patients with Down's syndrome (62). The gene encoding Cu,Zn-SOD is located on chromosome 21 in humans, and Down's syndrome is usually caused by trisomy of this gene, raising tissue Cu,Zn-SOD levels by about 50%. The limited data available at present are consistent with the view that the excess of Cu,Zn-SOD may contribute to at least some of the abnormalities in patients with Down's syndrome (62).
In addition to antioxidant defense enzymes, some low-molecular-mass free radical scavengers exist. Reduced glutathione can scavenge various free radicals directly, as well as being a substrate for GSHPX enzymes. a-Tocopherol is the most important free radical scavenger within membranes. Attack of reactive radicals, such as OH-, upon membranes can damage them by setting off a free radical chain reaction called lipid peroxidation. a-Tocopherol (a-TH) inhibits this by scavenging peroxyl radicals (Table 1), intermediates in the chain reaction.
a-TH + LOO--* aT. + LOOH [7] However, the tocopherol thereby becomes a radical, aT-. This illustrates a fundamental principle of free radical chemistry: when radicals react with nonradicals, new radicals are generated. Only when two radicals meet and join their unpaired electrons are the radicals lost (termination reactions). An example is the reaction of Owith NO (Equation 3).
Overall, antioxidant defenses seem to be approximately in balance with generation of oxygen-derived species in vivo. There appears to be no great reserve of antioxidant defenses in mammals, perhaps because, as pointed out previously, some oxygen-derived species perform useful metabolic roles.

Oxidative Stress: A Definition
Generation of ROS and the activity of antioxidant defenses appear more or less balanced in vivo. In fact, the balance may be slightly tipped in favor of the ROS so that there is continuous low-level oxidative damage in the human body. This creates a need for repair systems that can deal with oxidatively damaged molecules (Table 2). However, if a greater imbalance occurs in favor of the ROS, oxidative stress is said to result (19). Most aerobes can tolerate mild oxidative stress: indeed they often respond to it by inducing synthesis of extra antioxidant defenses. For example, if rats are gradually acclimatized to elevated 02, they can tolerate pure 02 for much longer than naive rats, apparently due to increased synthesis of antioxidant defenses in the lung (64,65). Other examples are the complex adaptive response of E. coli treated with low concentrations of H202 (66) and the activation of NF-KB in oxidatively stressed mammalian cells (33).
However, severe oxidative stress can cause cell damage and death. In mammalian cells, oxidative stress appears to cause increases in the levels of free Ca2+ (67) and free iron (68) within cells, e.g., by damaging proteins that normally keep these metal ions safely bound. Iron ion release can lead to OH-generation, which has been shown to occur within the nucleus of H202-treated cells (56). An excessive rise in intracellular free Ca2+ can also activate endonucleases and cause DNA fragmentation (67).
Hence, oxidative stress results in damage to DNA, proteins, lipids, and carbohydrates (67,68). The relative importance of damage to these different molecular targets in mediating cell injury or death depends upon what degree of oxidative stress occurs, by what mechanism it is imposed, for how long, and the nature of the system stressed. For example, lipid peroxidation appears to be an important consequence of oxidative stress in human atherosclerotic lesions (69). Several halogenated hydrocarbons (such as CC14 and bromobenzene) appear to exert some, or all, of their toxic effects by stimulating lipid peroxidation in vivo (7). However, for most other toxic agents causing oxidative stress, lipid peroxidation is not the major mechanism of primary cell injury: damage to proteins and DNA is usually more important (1,47,68). For example, it has often been assumed that lipids are a major target of damage by inhaled ozone, but proteins may be equally or more important (see below).

Causes of Oxidative Stress: Toxic Agents
Oxidative stress can be imposed in several ways. Thus, severe malnutrition can deprive humans of the minerals (e.g., Cu, Mn, Zn, Environmental Health Perspectives 8 Table 3. Why cigarette smoke can impose oxidative stress. (1) Smoke contains many free radicals, especially peroxyl radicals, that might attack biological molecules and deplete antioxidants, such as vitamin C and x-tocopherol.
(3) The tar phase of smoke contains hydroquinones. These are lipid-soluble and can redox-cycle to form°2and H20; They can enter cells and may even reach the nucleus to cause oxidative DNA damage. Some hydroquinones may release iron from the iron-storage protein ferritin in lung cells and respiratory tract lining fluids. (4) Smoking may irritate lung macrophages, activating them to make 0-.a (5) Smokers' lungs contain more neutrophils than the lungs of nonsmokers, and smoke might activate these cells to make 02-.a (6) Smokers often eat poorly and drink more alcohol than nonsmokers and may have a low intake of nutrient antioxidants.
"The effects of cigarette smoke on phagocytes are dose-related. Low levels may stimulate them, but high levels may poison them and so depress their activity.
Se) and vitamins (e.g., riboflavin-needed for the FAD cofactor of glutathione reductase, and cx-tocopherol-needed for antioxidant defense) (1). More usually, however, the stress is due to production of excess ROS.
Several drugs and toxins impose oxidative stress during their metabolism. Carbon tetrachloride is one example ( Table 1). Another is paraquat, a herbicide that causes lung damage in humans. Its metabolism within the lung leads to production of large amounts of 0;and H202 (1). Gas-phase cigarette smoke also imposes some oxidative stress. Some of the reasons for this are summarized in Table 3 (70,71).

Causes of Oxidative Stress: Disease and Tissue Injury
Does oxidative damage play a role in human disease? Many of the biologic consequences of excess radiation exposure may be due to OH--dependent damage to proteins, DNA, and lipids (6). Oxidative damage (resulting from exposure to elevated 02 in incubators) may account for damage to the retina of the eye (retinopathy Tissue damage by disease, trauma, poisons, and other causes usually lead to formation of increased amounts of putative "injury mediators," such as prostaglandins, leukotrienes, interleukins, interferons, and tumor necrosis factors (TNFs). All of these have at various times been suggested to play important roles in different human diseases. Currently, for example, there is much interest in the roles played by TNFa, NO-, and interleukins in adult respiratory distress syndrome and septic shock (8,72). ROS can be placed in the same category, i.e., tissue damage will usually lead to increased ROS formation and oxidative stress. Figure 1 summarizes some of the reasons for this. Indeed, in most human diseases, oxidative stress is a secondary phenomenon, a consequence of the tissue injury. That does not mean it is not important (1,72). For example, excess production of O°-202, and other species by phagocytes at sites of chronic inflammation can cause severe dam-* Increase in radical-generating enzymes (e.g. xanthine oxidase) and/or their substrates (e.g. hypoxanthine) -Activation of phagocytes -Activation of phospholipases, cyclooxygenases, and lipoxygenases I *-Dilution and destruction of antioxidants -Release of "free"metal ions from sequestered sites * Release of heme proteins (hemoglobin, myoglobin) -. Disruption of electron transport chains and increased electron leakage to form 02' Figure 1. Some of the mechanisms by which tissue damage can cause oxidative stress.
age. This seems to happen in the inflamed joints of patients with rheumatoid arthritis (72) and in the gut of patients with inflammatory bowel diseases (73). Tissue injury can release metal ions from their storage sites within cells, leading to OHgeneration (72,74). Thus, the main question about ROS in human disease is not "can we demonstrate oxidative stress?" but rather "does the oxidative stress that occurs make a significant contribution to disease activity?" The answer to the latter question appears to be "yes" in at least some cases, including atherosclerosis, rheumatoid arthritis, and inflammatory bowel disease (1,69,. However, it may well be "no" in many others. Elucidating the precise role played by free radicals has not been easy because they are difficult to measure, but the development of modern assay techniques is helping to solve this problem (72).

Causes of Oxidative Stress: Environmental Air Pollutants
The role of free radicals in the toxicity of°2 and of cigarette smoke (Table 3) has already been discussed, but oxidative damage is frequently suggested to be involved in the deleterious effects of 03 and NO2 (75,76). 03 is not a free radical, but it can oxidize many biological molecules directly and, in addition, it reacts slowly with water at alkaline pH to give OH- (77). It has also been suggested to produce singlet 02 when it reacts with biological molecules (78). The first biological fluids that come into contact with inhaled 03 are the respiratory tract lining fluids (RTLFs), that presumably serve to absorb and detoxify some of the inhaled 03 so as to lower the amount that enters the more vulnerable peripheral gas exchange regions of the lung. Some information is available about the antioxidants of these fluids (79,80) but the problems of sampling them (by the techniques of respiratory tract lavage) have hindered elucidation of their precise comparative antioxidant capabilities, since lavage itself produces considerable and variable dilution of RTLFs and some of their constituents may be oxidized during the procedures. In addition, the antioxidants present depend upon which part of the respiratory tract is being sampled (79), i.e., nasal passages, airways, bronchioles, alveoli. Often a mixed fluid is obtained.
By contrast, the antioxidant defenses of human plasma have been well characterized (53). To approach an understanding of how 03 might interact with a complex biological fluid, the reactions of 03 with freshly prepared human plasma have been studied.
Volume 102, Supplement 10, December 1994 Indeed, oxidant injury to the lung causes increased influx of plasma constituents, so that the lung lining fluids become more like plasma in composition. Uric acid and ascorbic acid were found to be the major plasma scavengers of 03 (81). Although it is often assumed that lipids are a major target of attack by 03, no evidence of substantial lipid damage by 03 was obtained (81), in keeping with other studies in the literature (82,83). Instead, oxidative protein damage was observed in 03-exposed plasma, as -SH group loss and protein carbonyl formation (84). However, it will be necessary to study human lung lining fluids to substantiate this conclusion, especially as Cueto et al. (85) found end products of lipid damage in lung lining fluids from rats after 03 exposure.
Since uric acid is a major constituent of upper airway lining fluid in humans (79), it could act as a "scrubber," decreasing the concentration of 03 in inhaled air so as to protect the more vulnerable alveolar regions of the lung (81). By contrast with the effect of 03, NO0 did not generate protein carbonyls in plasma, although nitration of aromatic amino acids took place and protein -SH groups were lost, presumably by direct reaction with NO (86). Again unlike 03, NO2 induced lipid peroxidation in plasma, presumably by the reactions L-H + NO -HN02 + L [8] L-+ 02-4 LOO [9] L-H + LOO --> L + LOOH [10] Initiation of peroxidation by NOQ (Equation 8) presumably sets up the autocatalytic chain reaction of lipid peroxidation (Equations 9,10) resulting in the accumulation of lipid peroxides (LOOH). Uric acid and ascorbate were found to be important antioxidants protecting plasma against oxidative damage by NO;, but the lipid-soluble antioxidants ct-tocopherol and ubiquinol probably also play an important protective role in limiting lipid peroxidation (86).
Hence, as summarized in Table 4, 03 and NO2 damage different molecular targets, which could be one reason why they might sometimes exert synergistic damaging effects in vivo. For comparison, Table 4 also includes the effects of cigarette smoke upon plasma. Ascorbic acid is rapidly oxidized and damage to both proteins (carbonyl formation, loss of -SH groups) and lipids (lipid peroxide formation) occurs. Ascorbic acid protects the plasma against lipid damage by ozone, but does not protect against protein damage (87,88).