Chapter 6 Mechanisms of cell injury by free radicals

Chapter 6 Mechanisms of cell injury by free radicals

Chapter 6 Mechanisms of Cell Injury by Free Radicals JOE M. MCCORD Introduction Free Radicals and Active Oxygen Species: Why is Oxygen So Eager to ...

815KB Sizes 0 Downloads 57 Views

Chapter 6

Mechanisms of Cell Injury by Free Radicals

JOE M. MCCORD

Introduction Free Radicals and Active Oxygen Species: Why is Oxygen So Eager to Form Free Radicals? How Cells Deal with Oxygen Biological Sources of Free Radical Production Superoxide and Bactericidal Action by Phagocytes Superoxide Production by Mitochondria Superoxide Production by Xanthine Oxidase Other Sources of Active Oxygen Production Why So Many Diseases Involve Free Radical Production How Does Superoxide Radical Cause Tissue Injury Iron Exacerbates Free Radical Injury and Most of Us are Iron-Overloaded The Oxidant-Antioxidant Balance Free Radicals and Vascular Tone Conclusion

Principles of Medical Biology, Volume 13 Cell Injury, pages 197-211. Copyright 9 1998 by JAI Press Inc. All rights of reproduction in any form reserved. ISBN: 1-55938-818-8 197

198 198 200 200 200 202 202 203 203 204 205 209 209 210

198

JOE M. McCORD

INTRODUCTION The discovery in 1969 of the enzyme superoxide dismutase (McCord and Fridovich, 1969) caused a certain amount of bewilderment in biological circles. The enzyme catalyzed the seemingly unlikely radical-radical annihilation (or disproportionation) reaction of the superoxide radical (O ~-): o-

o-

02 +02 + 2 H §

H202+O 2

Did organisms really possess an enzyme for the purpose of getting rid of free radicals, or was this activity incidental, some sort of curious test-tube artifact? Free radicals had previously been in the domains of radiation chemists, polymer chemists, and plasma physicists, but not of biologists. The existence of a free-radicalscavenging enzyme implied, of course, that there were significant biological sources of free radical production (at least of the superoxide radical). Today it is known that free-radical production plays roles in numerous diseases. The vast majority of these diseases involve two basic and recurring mechanisms of free-radical production: activated inflammatory cells and injury resulting from ischemia and reperfusion. Equally important, we have finally come to the realization that the superoxide radical has a dual personality, a good side and a bad side. Recent studies suggest that cellular redox status, determined in part by rates of superoxide generation and scavenging, may play important roles in regulating cellular proliferation (Murrell et al., 1990; Burdon, 1995) and apoptosis (Buttke and Sandstrom, 1994). Overproduction of the radical has long been associated with increased lipid peroxidation, but we now find that over-scavenging of the radical can paradoxically lead to the same end (Nelson et al., 1994). This is because free radicals can initiate chain reactions, and they can also terminate them. Hence, the relative rates of free-radical production, reaction, and scavenging produce a rather delicate and precarious balance in the healthy cell.

FREE RADICALS AND ACTIVE OXYGEN SPECIES: WHY IS OXYGEN SO EAGER T O F O R M FREE RADICALS? Chemically, a free radical is any molecule containing a single, unpaired electron. Because a covalent bond is formed when a pair of electrons occupies a single molecular orbital, free radicals may be viewed as molecular fragments formed by homolytic cleavage of a bond. (Usually, transition-metal ions are not considered free radicals, although by technical definition they are.) The unpaired electron imparts paramagnetic character to the radical, which displays a characteristic absorption spectrum in a magnetic field, which is detectable by electron paramagnetic resonance spectroscopy (EPR). Molecular oxygen is somewhat of an electronic oddity

Mechanisms of Cell Injury by Free Radicals

199

in that its ground state is a diradical, with two unpaired electrons of like spin. This unconventional distribution of electrons makes it impossible for oxygen to accept a spin-matched pair of electrons, as badly as it may want them, until one of its unpaired electrons undergoes a spontaneous spin reversal to make pairing possible. At ordinary collisional frequency, the period of contact is too brief, imposing a kinetic barrier to most oxidative reactions. It is this kinetic barrier that saves us from reacting explosively with an atmosphere of huge thermodynamic potential. It is the same kinetic restriction that renders oxygen an ideal terminal electron acceptor for biological systems. Enzymes with a binding site for oxygen can keep it in contact with its reductant for a much longer time than would occur by simple collision, thereby overcoming the kinetic reluctance to react and rendering available the huge amounts of energy to be derived from giving electrons to molecular oxygen. Our relationship with oxygen is still an uneasy one. Oxygen has great affinity for four more electrons. (The process of taking away electrons is called "oxidation" because oxygen does it so well. The substance receiving electrons becomes "reduced.") This complete reduction of oxygen breaks the bonds holding the two nuclei together and after picking up protons from the solvent, results in the formation of two molecules of water. Occasionally, under normal biological conditions, oxygen does manage to steal away electrons from other molecules by nonenzymatic autoxidations. Because it cannot accommodate a spin-matched pair, oxygen must acquire electrons one at a time. This separation of electron pairs results in free radical formation. The one-electron reduction product of oxygen is the superoxide radical, O 2- (see Figure 1). If two electrons are transferred, the product is hydrogen +e"

+e-

02 ---. 0 2

+e"

+e

+2.=* H202 __~ HO"

2 H20

+OH NADPH oxidase, xanthine oxidase, mitochondria J

flavoprotein oxidases, dismutation of superoxide

Fenton chemistry (superoxide, hydrogen peroxide, and iron) ....

I

cytochrome c oxidase

J

Figure 1. The one-, two-, three-, and four-electron reduction products of oxygen. Products may be formed by sequential transfer of single electrons or by concerted transfer of multiple electrons as catalyzed by cytochrome c oxidase. Also shown are representative physiological sources for each reactive oxygen intermediate.

200

JOE M. McCORD

peroxide H202, which is not a radical. It is nonetheless still hungry for two more electrons, making hydrogen peroxide a cytotoxic oxidant. Ferrous iron is capable of transferring a third electron to hydrogen peroxide, causing lysis of the O-O bond. One fragment is reduced to water. The other fragment is the hydroxyl free radical HO ~ a very potent oxidant. It can initiate lipid peroxidation, cause DNA strand breaks, and indiscriminately oxidize virtually any organic molecule. This random action actually works in our favor, because most of the targets it strikes are expendable. Reactivity and toxicity are not synonymous, as will be discussed later. The family of reactive intermediates resulting from the incomplete reduction of oxygen includes superoxide radical, hydrogen peroxide, and hydroxyl radical. It is not correct to refer to this group as oxygen-derived free radicals, because one member is not a radical. Accordingly, several terms are now used to refer to this family. The most common term is reactive oxygen metabolites (ROM), active oxygen (AO), or variations thereon. Occasionally, the terms are extended to include electronically excited oxygen (singlet oxygen) and even hypochlorous acid, which is produced from hydrogen peroxide by neutrophil myeloperoxidase.

How Cells Deal With Oxygen Oxygen is handled almost entirely by a very few so-called "professional oxidases." These enzymes have mastered the trick of feeding electrons to oxygen without creating the undesirable, partially reduced intermediates described above. Cytochrome oxidase (or the a/a 3 complex) is the terminal electron carrier of our mitochondria. This enzyme single-handedly deals with about 98% of the oxygen that we metabolize, reducing it to water with no detectable production of intermediates. Other oxidases (especially flavoproteins) may reduce oxygen to hydrogen peroxide (which is detoxified by catalase) without producing any superoxide. It appears to be the enzymes that are designed to reduce pyridine nucleotides (the "nonprofessional oxidases") that tend to make small amounts of superoxide when they interact with oxygen. Most of the superoxide produced by a healthy cell results from tiny amounts of "leakage" or "short-circuiting" of electrons, analogous to breaks in the insulation of an electronic circuit. Sick cells, like damaged circuit boards, have increased leakage.

BIOLOGICAL SOURCES OF FREE RADICAL P R O D U C T I O N Superoxide and Bactericidal Action by Phagocytes Nature has a knack for putting things to constructive uses. As free radical biologists were acknowledging the unavoidable production of superoxide and the protective roles of superoxide dismutases, an insightful observation was made by Dr. Bernard Babior. It had long been known that neutrophils display a burst of cyanide-

Mechanisms of Cell Injury by Free Radicals

201

insensitive (i.e., nonmitochondrial) oxygen consumption when they encounter and kill microorganisms. It was further known that this resulted in producing hydrogen peroxide, which can kill microorganisms. Babior thought superoxide radical might be a likely candidate to participate in bactericidal action and showed that neutrophils stimulated to engulf latex particles did indeed produce this free radical. The bactericidal roles of the free radical and its metabolites H202 and HOC1 are now well understood (Babior, 1978). Much subsequent work has characterized a multicomponent NADPH oxidase located in the plasma membrane of the neutrophil. This enzyme may be the only enzyme in the body that produces superoxide by design rather than by accident. The genetic inability to produce superoxide in this instance causes the life-threatening condition known as chronic granulomatous disease (or CGD). Neutrophils from such individuals have seriously impaired ability to kill microorganisms that have been ingested, leading to multiple recurrent local infections and often to septicemia and death at an early age. True symbiosis is relatively rare. Most life forms do not tolerate being invaded by other life forms. Such a challenge usually precipitates a fight to the death of one party or the other. In higher organisms, the neutrophil is the foot soldier of this war for self-preservation. It is programmed in the philosophy of the Old West: "Shoot first and ask questions later." So much is at risk that failure to carry out its mission is unacceptable. It has to be prepared to fire away at any antigen it does not recognize, assuming dire consequences if it allows the foreigner to survive. Hence, it is programmed for overkill, not for caution. We pay a substantial price for this ultraconservative policy. The "antibiotics" in the neutrophil's arsenal (O2-, H202, and HOC1) have the broadest possible spectrum. They are not the selective silver bullets of our modern-day pharmacopoeia. Therefore, the battlefield becomes littered with casualties. Host tissues are caught in the cross fire and succumb to the oxidant attack, along with the neutrophils themselves. Infected tissues display the cardinal signs of inflammation (redness, heat, swelling, pain, and loss of function) not necessarily as a result of the invading microbe but largely due to the war waged by the attacking neutrophils and the unavoidable damage to host tissues. Because all neutrophils must pass through the lung on every round-trip through the circulation, because the lung is structurally rather delicate and easily injured, and because loss of pulmonary function for even brief periods results in death, the lung is a very vulnerable target for neutrophils, whether activated systemically or locally. It has been hypothesized that various events known to incite the adult respiratory distress syndrome (ARDS) cause abnormal accumulation of adherent neutrophils in the vasculature of the lung, possibly via complement-mediated mechanisms, and that the tissue injury resulting from neutrophil stimulation and release of superoxide radical and related oxidants contributes to the development of ARDS. If a pathogenic infection is squelched, the cost of the warfare may be justified. Often, however, the war is waged at a perceived but nonexistent or nonthreatening enemy. This is the case in autoimmune diseases and allergies. When an antibody is produced against any antigen, that component is marked by the formation of an im-

202

JOE M. McCORD

mune complex and is designated as a target for the neutrophil. If the antigen is a soluble component, then a circulating immune complex is formed. This may activate neutrophils as above, or the immune complex may localize in the lung, kidney, or other tissues, attracting and activating the neutrophils and instigating tissue injury. This mechanism is thought to be active in systemic lupus erythematosus and the associated nephritis, in Crohn's disease (ulcerative colitis or inflammatory bowel disease), and in asthma. Multiple sclerosis is now being viewed as a possible autoimmune disease with evidence of active oxygen-mediated injury, as is rheumatoid arthritis. In other types of arthritis, neutrophil activation may result from physical trauma to the joint tissues (osteoarthritis or degenerative joint disease) or from attempts to phagocytose urate crystals (gout).

Superoxide Production by Mitochondria The production of partially reduced oxygen species by the mitochondrion has been known for nearly two decades. The mechanisms of production of active oxygen species, however, have been examined only more recently and are still debated. The rate of superoxide production by the mitochondrion increases when the concentration of oxygen is increased or the respiratory chain becomes largely reduced. The first process, for example, takes place in the lungs of animals exposed to high oxygen concentrations. The second process occurs when coupled mitochondria consume oxygen in state 4 (that is, in the absence of ADP and in the presence of substrate and oxygen), when the electron flow is limited by the lack of substrate to phosphorylate. Mitochondria produce superoxide anions at two sites in the electron transport chain. The first site is the ubiquinone to cytochrome c~ step, which passes through the intermediate ubisemiquinone. Ubisemiquinone is capable of reducing oxygen to superoxide, which dismutates spontaneously to form hydrogen peroxide. The second site of superoxide anion formation is the NADH-dehydrogenase. Several studies have demonstrated that heart mitochondria become uncoupled upon exposure of the organ to ischemia. It has also been observed that after ischemia, mitochondrial electron transport is specifically inhibited between NADH-dehydrogenase and ubiquinone. This inhibition is associated with increased rate of formation of hydrogen peroxide by NADH-dehydrogenase. There is evidence that the inhibition of the mitochondrial respiratory chain may be caused by the uptake of calcium. It has been shown that mitochondria accumulate calcium electrogenically, using the H § gradient for this purpose instead of producing ATP, even in the presence of ADP. It is possible that calcium overload of mitochondria may activate phospholipases or proteases, either of which may cause uncoupling and inhibition of respiration.

Superoxide Production by Xanthine Oxidase Xanthine oxidase exists in vivo primarily as a dehydrogenase, which uses NAD § as an electron acceptor during the oxidation of xanthine and hypoxanthine. In tis-

Mechanisms of Cell Injury by Free Radicals

203

sues containing abundant xanthine dehydrogenase, this enzyme becomes an important source of superoxide upon tissue reoxygenation after ischemia or after extreme hypotension (as in hemorrhagic shock). During ischemia there is massive breakdown of the adenine nucleotide pool due to the low energy status of the tissue. Adenosine is converted to inosine and then to hypoxanthine, which accumulates in abundance. Although about 10% of any tissue's xanthine dehydrogenase exists as an oxygen-utilizing superoxide-producing xanthine oxidase, ischemia-induced proteolytic conversion appears to result in even more of the radical-producing form of the enzyme. Hence, after a period of ischemia, there is xanthine oxidase (preexisting or newly converted) and an abundance of its purine substrate hypoxanthine. At reperfusion the remaining substrate, molecular oxygen, reenters the tissue, and a burst of superoxide production ensues. Reperfused organs are dramatically protected by inhibitors of xanthine oxidase or by superoxide dismutase. There is great tissue and species variability in the content of xanthine dehydrogenase/oxidase. Both the rat and the dog contain at least several orders of magnitude higher amounts of xanthine oxidase in their hearts than rabbit, pig, or human. Enzyme-inhibiting doses of allopurinol do not protect the rabbit heart, although protection is still afforded by superoxide dismutase. This observation implies the existence of a xanthine oxidase independent mechanism of superoxide production, most likely resulting from ischemic injury to the mitochondria, as described above.

Other Sources of Active Oxygen Production While the three sources described above probably represent the major sources of reduced oxygen metabolites, several other sources have been suggested. Arachidonic acid metabolism via PGH synthase and lipoxygenase appears to generate superoxide radical. It has been proposed that the autoxidation of catecholamines (which generates superoxide) causes myocardial damage during ischemia and reperfusion. Hemoproteins, such as myoglobin, may participate in active oxygen dependent injury.

WHY SO MANY DISEASES INVOLVE FREE RADICAL PRODUCTION The quantitatively major sources of free-radical production, ischemia/reperfusion, inflammation, and mitochondrial injury, are basic and common components of a vast number of diseases. Furthermore, one source of free-radical production often leads to additional sources. In vivo, ischemia/reperfusion injury invariably involves subsequent inflammatory injury. In fact, antineutrophil measures, such as administration of an antiadherent monoclonal antibody, provide nearly complete suppression of injury to the postischemic feline intestine. If protection in this model may be provided by inhibitors of xanthine oxidase, by superoxide dismutase, or by anti-

204

IOE M. McCORD

neutrophil measures, how can all these studies be reconciled? The answer seems to lie in the temporal sequence of the events. Relatively small amounts of O~- may be produced in the reperfused tissue by xanthine oxidase, immediately upon reperfusion. The amount of superoxide initially formed may be insufficient to cause massive tissue injury but may initiate an infiltration of neutrophils by activating a superoxide-dependent chemoattractant. This chemoattractant is activated in even greater amount with the arrival of the first wave of neutrophils, which upon sti~ulation by components leaking from the damaged tissue, generate still more superoxide. The factor thus serves to amplify and maintain the process of chemotaxis. Most of the tissue injury results from the oxidants generated by huge numbers of neutrophils invading the tissue. Then, xanthine oxidase inhibitors block the initial influx of neutrophils. Superoxide dismutase inhibits the initial chemotaxis and the amplified and maintenance phases of chemotaxis. Antiadhesive antibodies prevent neutrophils from sticking to the vessel walls, so the cells cannot leave the circulation to infiltrate a tissue.

HOW DOES SUPEROXIDE RADICAL CAUSE TISSUE INJURY? Superoxide radical is not a highly reactive free radical like hydroxyl radical. In fact, its lack of reactivity caused some to question the need for the family of protective enzymes known as the superoxide dismutases. Now, however, it is clear that superoxide production beyond a certain point becomes toxic to the cell. Superoxide reacts avidly with certain targets. The vast majority of enzymes are unaffected by superoxide. Table 1 lists some enzymes that are quite sensitive to being inactivated by superoxide. This is not necessarily an exhaustive list. (The list becomes much longer if one includes those enzymes that are inactivated by secondarily generated radicals, including hydroxyl. In fact, any enzyme in pure solution will be inactivated by exposure to a sufficient quantity of hydroxyl radical, but that does not mean it would be inactivated by hydroxyl radical inside a cell, as will be discussed below.) Scanning the list, one finds some fairly important enzymes. Creatine phosphokinase is vital to high-energy phosphate metabolism in heart and muscle. Catalase and glutathione peroxidase are particularly notable because they are sister antioxidant enzymes to superoxide dismutase. Superoxide dismutase, then, protects these two antioxidant enzymes from inactivation. Ironically one of the few things that will inactivate superoxide dismutase is the substrate for these two enzymes, hydrogen peroxide. Thus, there is a system of mutual protection among the antioxidant enzymes. Certain families of enzymes with common structural similarities may emerge as particularly sensitive targets: enzymes relying on an essential sulfhydryl group (e.g., creatine phosphokinase and papain) or enzymes containing a 4-iron, 4-sulfur center (e.g., 6-phosphogluconate dehydratase and aconitase). The selective reactivity of superoxide may lead to much greater mole-formole toxicity than for the hydroxyl radical. The latter is widely accepted as being

Mechanisms of Cell Injury by Free Radicals

205

Table 1. A List of Enzymes Known to Be Inactivated by Reaction with the Superoxide Radical. Enzyme Source Catalase Mammalian liver 3-or-Hydroxysteroid dehydrogenase Pseudomonastestosteroni Glyceraldehyde-3-phosphate dehydrogenase Mammalian muscle Papain Plant o~,13-Dihydroxyisovaleratedehydratase Escherichiacoli Ornithine decarboxylase Mammalian heart Glutathione peroxidase Mammalian Hydrogen-NAD§ oxidoreductase Alcaligeneseutrophus Myofibrillar ATPase Mammalian heart Adenylate cyclase Mammalian brain ....

Creatine phosphokinase

Mammalian heart

6-Phosphogluconate dehydratase

Bacterial

Aconitase

Bacterial

Glutamine synthetase

Mammalian brain

NADH-ubiquinone oxidoreductase

Bovine mitochonclria

ATPase

Bovine mitochondria

Succinate ubiquinone oxidoreductase

Bovine mitochondria

Alcohol dehydrogenase Ca2§

Yeast Mammalian smooth muscle encloplasmic reticulum

extremely reactive, but not at all selective. Hence, the vast majority of hydroxyl radicals will react with cellular targets than are nonvital and expendable.

IRON EXACERBATES FREE RADICAL INJURY AND MOST OF US ARE IRON-OVERLOADED Even before it was realized that iron is a necessary trace element essential for life itself, there was a natural tendency to z,ssociate iron with strength. The ancient Greeks concocted potions of iron dissolved in vinegar, hoping to acquire the properties of the element. After its recognition as an essential nutrient, the natural but incorrect assumption followed that more must be better. This assumption persists today. Physicians and scientists have been no more immune from this line of reasoning than the rest of society. Accordingly, for several decades we have been sold on the idea of "iron-fortified" foods, in many cases to our eventual detriment.

206

JOE M. McCORD

When it was realized that biological systems can produce free radicals and other potent oxidants, the natural tendency of conventional wisdom was to label the oxidants as "bad" and the antioxidants as "good." Things are rarely so simple. Now, after two decades of contemplation and experimentation, there is a growing recognition that a balance between oxidants and antioxidants is a more realistic depiction of the relationship. There were many early clues, most of which were ignored or dismissed due to their circumstantial nature, or possibly just because they flew in the face of conventional wisdom. With the recognition that life must carefully juggle oxidants against antioxidants has come the recognition that iron, the most common redox-active transition element present in biological systems, is an active participant in this precarious balancing act. From a physiological or clinical perspective, the important point to appreciate is that iron may seriously exacerbate any oxidative stress. Soon after the discovery that biological systems can produce superoxide, evidence suggested that in some cases superoxide radical was not acting alone. It appeared to be collaborating with hydrogen peroxide (the product of its dismutation) to produce a species with much greater oxidizing potential than either of the coconspirators. This new species was presumed to be the hydroxyl radical (HO.), one of the most potent oxidants known. It was thought to be generated via a reaction first proposed in 1934 by Haber and Weiss: o0 2 "~- H 2 0

2 --9'

0 2 -~- H O "

+OH

(1)

The rate constant for this reaction is so slow as to preclude biological relevance. It was found however, that iron catalyzes the reaction: O2- + Fe .3 ~ 02 + Fe .2

(2)

Fe*2+ H20 2 ~ Fe § + HO" + O H -

(3)

e-

O~ + H 2 0 ~ ~ O~ +HO" +OH

(1)

This combination of reactions is now commonly referred to as the ironcatalyzed Haber-Weiss reaction or as superoxide-driven Fenton chemistry, as reaction (3) had been proposed much earlier by Fenton. It also seems proper to refer to it simply as Haber-Weiss chemistry, because these scientists postulated all three reactions. Higher organisms are particularly fastidious about handling iron. There is never an appreciable amount of "flee" or loosely chelated iron in the healthy state. It is transported in the ferric state, bound to transferrin in a complex especially difficult to reduce. Likewise it is stored in the ferric state by ferritin, a protein found in virtually all tissues and in plasma. Importantly, the superoxide radical is capable of reducing ferritin-bound iron to the ferrousstate, whereupon it is released:

Mechanisms of Ceil Injury by Free Radicals O 2- + Ferritin - Fe § ~ O 2

207 "~" Ferritin

+ Fe +2

(4)

It is this iron, liberated by the pathological production of superoxide, that is now free to catalyze Haber-Weiss chemistry. The hydroxyl radical produced can initiate lipid peroxidation, a free radical chain reaction leading to loss of membrane structure and function. The hydroxyl radical can also attack structural macromolecules and cause DNA breaks. The liberated iron can also cause the reductive lysis of the oxygen-oxygen bond in a preexisting lipid hydroperoxide molecule, giving rise to a lipid alkoxyl radical (LO') which may then serve as an initiating radical for lipid peroxidation: Initiation: Propagation:

Fe 2+ + LOOH ~ Fe 3+ + LO" + OHLO" + LH --~ LOH + L" L" + 0 2 ~ LOO" LOO" + LH ~ LOOH + L"

(5) (6)

(7) (8)

If "liberated" iron can seriously exacerbate the component of tissue injury due to free radical production, the question arises as to whether iron status from a nutritional standpoint might be a predisposing factor in all of the disease states discussed above. American males accumulate total body iron stores almost linearly after puberty, commonly reaching iron stores of two grams or more. Women are protected until the age of menopause, after which time their iron accumulation parallels that of men. The body absorbs dietary iron with an efficiency of about 10%, but it has no mechanism to get rid of excess iron. Hearts from iron-loaded animals suffer substantially more injury when subjected to ischemia/reperfusion than hearts from normal animals. A prospective human study has concluded that iron sufficiency is associated with hypertension and excess risk of heart attack (Salonen et al., 1992). In a study of patients with small cell carcinoma of the lung, it was found that those patients with the lowest serum ferritin levels at the time of diagnosis had significantly longer survival times. The probable sequence of events leading from the pathological overproduction of active oxygen products (O ~- and H202) to cell injury and death is summarized in Figure 2. The first lines of defense against these agents are the antioxidant enzymes superoxide dismutase, glutathione peroxidase, and catalase. If these enzymes are overwhelmed or otherwise compromised, iron is liberated from tissue stores of ferritin. This iron can be redox-cycled by superoxide radical and can initiate lipid peroxidation directly (by reducing preexisting lipid peroxides to create initiating alkoxy radicals) or indirectly (by catalyzing the production of hydroxyl radical, another initiating radical). Once lipid peroxidation is in progress, it spreads like a brush fire through cell membranes. The antioxidant vitamin E is a lipid-soluble vitamin that inserts itself into the membrane to serve as a "firebreak." When a propagating radical hits a vitamin E molecule (VitE-H2), a very stable radical (VitE-H') is formed that does not participate in continuing the chain reaction. Rather, it mi-

208

JOE M. McCORD

Disease Processes 02- H 2 0 2 Superoxide dismutase . . . / ......."~" ---I............ Glutathione

Iron Stores

Ferritin

e'y"

(

~

++

J

peroxidase

" ' " ....... Catalase

O"

Lipid peroxidation

iiiii..iiiii.)....

Phospholipid hydroperoxide glutathione peroxidase "'"

Cell Injury and Death Figure 2. The probable sequence of events leading from overproduction of active oxygen species to cell injury and death. Antioxidant enzymes and vitamins are shown in italic, and their sites of intervention are indicated by dotted lines. grates to the aqueous interface where it is reduced by reacting with the watersoluble vitamin C to regenerate functional vitamin E. Vitamin C is consumed in the process, forming its oxidized product, dehydroascorbate: Termination: LOO" + V i t E - H 2 -'> LOOH + V i t E - H" 2 VitE - H" + VitC ~ 2 VitE - H 2 + Dehydroascorbate

(9) (10)

A certain amount of damage control is provided by the selenium-dependent antioxidant enzymes glutathione peroxidase and phospholipid hydroperoxide glutathione peroxidase. Both utilize reduced glutathione (GSH) to eliminate hydroperoxides from the stressed cell. If allowed to remain, these compounds may react with ferrous iron to set off new rounds of lipid peroxidation (see Eq. (5)). The former enzyme acts on fatty acid hydroperoxides which have been released from membrane phospholipids by the actions of phospholipases. The latter enzyme acts on plaospholipid hydroperoxides still involved in membrane structures. In either case, the potentially reactive hydroperoxide is reduced to an unreactive alcohol: LOOH + 2 GSH ~ LOH + H 2 0 + G S S G

(11)

Mechanisms of Cell Injury by Free Radicals

209

If lipid peroxidation gets out of hand and cannot be contained, sufficient loss of membrane structure and function results in irreversible injury and death.

THE OXIDANT-ANTIOXIDANT BALANCE In healthy organisms, a relatively delicate balance is maintained among oxidants (such as active oxygen species and transition metals) and antioxidants (such as superoxide dismutase, catalase, vitamins E, C, and A). Under pathological conditions the balance may be tilted toward the oxidative side, as outlined above, or more rarely toward the reductive side. The end result is uncontrolled and potentially lethal lipid peroxidation in either case. Patients with Down syndrome have three copies of the gene for cytosolic superoxide dismutase and express 150% of the normal amount of this enzyme. Their cells show an increased tendency to undergo lipid peroxidation, presumably because they have steady-state concentrations of superoxide radical too low to help annihilate the lipid dioxyl radicals that propagate lipid peroxidation (Nelson et al., 1994). The antioxidant vitamins E (lipid soluble) and C (water soluble) collaborate to terminate lipid peroxidation chain reactions. A recent summary of a number of European population studies underscores the protective role of the antioxidant vitamins in heart disease (Gey et al., 1991). This study correlated mortality from ischemic heart disease with serum vitamin E levels and with two factors considered to be major risk factors, hypertension and serum cholesterol. The correlation was stronger versus vitamin E than with either of the other two factors. The higher the plasma vitamin E level, the lower the mortality rate from heart disease. Iron, of course, exacerbates any oxidative imbalance, as described above.

FREE RADICALS AND VASCULAR TONE A recent development in free-radical biology has been the recognition of the role of nitric oxide (NO.) as an endothelium-derived relaxing factor. Nitric oxide is a colorless paramagnetic gas generated by the action of nitric oxide synthase on the amino acid L-arginine. The nitric oxide radical has a half-life of only a few seconds under physiological conditions, decomposing quickly via an oxygen dependent disproportionation to ultimately form nitrite and nitrate: 2 NO" + 0 2 ~ 2 N O 2 2 N O 2 + H 2 0 - ~ NO 2 + N O ; +2H +

(12) (13)

The most important physiological action of nitric oxide appears to be the stimulation of soluble guanylate cyclase, the enzyme responsible for producing cyclic GMP (cGMP). This provides the basis for the molecule's ability to cause relaxation of vascular smooth muscle cells (or vasodilation). Because nitric oxide is an oxidiz-

210

JOE M. McCORD

ing free radical, it has certain reactivities in common with superoxide. Specifically, it can inactivate certain enzymes listed in Table 1, especially those containing the 4-iron, 4-sulfur center, such as aconitase and the electron transport enzymes NADH-ubiquinone oxidoreductase and succinate ubiquinone oxidoreductase. Superoxide dismutase has vasodilatory activity, presumably by preventing an annihilation reaction between nitric oxide and superoxide:

NO" +O~- ~ ONOO-

(14)

The product of this reaction, peroxynitrite, is itself a potent oxidant. In many systems it exhibits properties similar to the hydroxyl radical. Thus, the potential interactions between nitric oxide and superoxide are rather complex in nature, to the extent that it is difficult to predict whether any particular reaction may, ultimately, be protective or destructive to the organism.

CONCLUSION A number of factors contribute to a nearly ubiquitous role for the generation ofreacfive oxygen species under pathological circumstances. Oxygen is prone to radical generation due to a quirk in its electronic structure. To generate energy, virtually all cells must deal with the chemically difficult reduction of oxygen. The fact that neutrophils have learned to put cytotoxic active oxygen to a constructive use is a mixed blessing. Although it saves us from infection, it has enormous potential for host tissue destruction when generated inappropriately, as in autoimmunity. Even though all cells contain defenses to combat the onslaught of reactive oxygen species, the delicate balance may become skewed during periods of oxidative stress, resulting in cell injury or death.

REFERENCES Babior, B.M. (1978). Oxygen-dependent microbial killing by phagocytes. N. Engl. J. Med. 298, 659-668, and 721-725. Burdon, R.H. (1995). Superoxideand hydrogenperoxide in relation to mammaliancell proliferation. Free Radical Biol. Med. 18, 775-794. Buttke, T.M. and Sandstrom,P.A. (1994).Oxidativestress as a mediatorof apoptosis. Immunol.Today 15, 7-10. Gey, K.F.,Puska, P., Jordan,P., and Moser,U.K. (1991).InversecorrelationbetweenplasmavitaminE and mortalityfromischemicheart disease in cross-culturalepidemiology.Am. J. Clin. Nutr. 53 (Suppl.), 326S-334S. McCord, J.M. and Fridovich,I. (1969). Superoxidedismutase: An enzymicfunctionfor erythrocuprein (hemocuprein). J. Biol. Chem. 244, 6049-6055. Murrell, G.A.C., Francis, M.J.O., and Bromley,L. (1990). Modulation of fibroblast proliferationby oxygen free-radicals. Biochem. J. 265, 659-665.

Mechanisms of Cell Injury by Free Radicals

211

Nelson, S.K., Bose, S.K., and McCord, J.M. (1994). The toxicity of high-dose superoxide dismutase suggests that superoxide can both initiate and terminate lipid peroxidation in the reperfused heart. Free Radical Biol. Med. 16, 195-200. Salonen, J.T., Nyyssonen, K., Korpela, H., Tuomilehto, J., Seppanen, R., and Salonen, R. (1992). High stored iron levels are associated with excess risk of myocardial infarction in eastern Finnish men. Circulation 86, 803-811.

RECOMMENDED READINGS Forman, H.J. and Cardenas, E. (1997). Oxidative Stress and Signal Transduction. Chapman and Hall, London. Meneghini, R. (1997). Iron homerstasis, oxidative stress and DNA damage. Free Rad. Biol. Med. 23, 783-792. Scandalios, J.G. (1997). Oxidative Stress and the Molecular Biology of Antioxidant Defenses. Cold Spring Harbor Laboratory Press, Plainview, N.Y. Sies, H. (1997). Antioxidants in Disease Mechanisms and Therapy. In: Advances in Pharmacology. Vol. 38, Academic Press, San Diego.