Current Concepts of Cerebrovascular Disease and Stroke. Free Radicals in Central Nervous System Ischemia

Transcription

1 1086 Current Concepts of Cerebrovascular Disease and Stroke Free Radicals in Central Nervous System Ischemia J.W. Schmidley, MD Downloaded from by on November 22, 2018 Free radicals have been implicated in a wide variety of diseases and in the toxic and therapeutic effects of antineoplastic drugs and radiation, the deleterious consequences of environmental pollutants, and "degenerative" processes such as aging and Parkinson's disease. 1-2 Their involvement in myocardial and intestinal as well as central nervous system (CNS) ischemia is under intensive study. 3 A free radical is any molecule, atom, or group of atoms with an unpaired electron in its outermost orbital. Since covalent chemical bonds usually consist of a pair of electrons sharing an orbital, free radicals can be thought of as molecules with an "open" or "half bond, which accounts for their extreme reactivity. 3 Free radical species of potential importance in cerebral ischemia include superoxide (O 2 ~-) and hydroxyl (OH-). [By convention, the single unpaired electron in a free radical is represented by a dot.] In the acidic conditions of ischemic brain, O 2 ~- is probably protonated (HO 2 ~-). The OH- radical is the more reactive and more toxic of these two molecules. Hydrogen peroxide (H 2 O 2 ), while not a free radical per se, has the potential to generate OH- radicals in reactions with O 2 ~-, catalyzed by iron (or other transition metals): O 2 ~+H 2 O 2 -»O 2 +OH~+OH-. In order to function as a catalyst for this reaction, the Fe must not be bound to proteins. Since unlike most extracellular fluids the cerebrospinal fluid (CSF) has low concentrations of Fe-binding proteins, iron released from damaged brain cells is more likely to be readily available to catalyze the generation of OH-. Fe can also donate electrons to H 2 O 2 to form OH-: H 2 O 2 +Fe 2+^Fe 3+ +OH-+0H". Because H 2 O 2 is nonpolar, it readily crosses membranes unlike O 2 ~\ Free radicals are produced in small amounts by normal cellular processes. The mitochondrial electron transport system is designed to add four electrons to O 2, reducing it to H 2 O and avoiding the reactive species produced by single electron reduction of O 2. However, "leaks" in mitochondrial electron transport allow O 2 to accept single electrons, forming O 2 ~\ Free radicals are produced in the reactions catalyzed by prostaglandin hydroperoxi- From the Department of Neurology, Case Western Reserve University, School of Medicine, Cleveland, Ohio. Reprinted from Current Concepts of Cerebrovascular Disease and Stroke 1990;25:7-12. dase and are byproducts of the normal or pathologic function of several other enzymes. They are also produced in cells by the auto-oxidation of various small molecules including catecholamines and by the microsomal cytochrome P-450 reductase system. 4 The features of free radical chemistry central to any potential role in cerebral ischemia are their extreme reactivity and their tendency to initiate and participate in chain reactions. When a free radical with its lone electron reacts with another molecule, another free radical must be produced (Figure 1). This free radical in turn can react with another molecule and so on until the chain of reactions is terminated either by the random collision of two free radicals to form a molecule with a stable bond or by one of the cellular defense mechanisms discussed below (references 3-8 should be consulted for details of free radical chemistry). Free radicals can react with and damage proteins, nucleic acids, lipids, and other classes of molecules such as the extracellular matrix glycosaminoglycans (e.g., hyaluronic acid). The sulfur-containing amino acids and the polyunsaturated fatty acids are particularly vulnerable. Because the latter are found in high concentrations in the CNS, most research on the role of free radicals in cerebral ischemia has concentrated on these molecules. Fatty acids are most susceptible to free radical attack at alpha methylene carbons, those adjacent to carbon-carbon double bonds (Figure 1). Polyunsaturated fatty acids with several double bonds per molecule are therefore particularly liable to free radical damage. 5 Unicellular and multicellular organisms have a variety of defenses against free radicals, among them the low molecular weight "scavengers" such as alphatocopherol and ascorbate. Alpha-tocopherol (vitamin E) is lipid soluble and therefore easily crosses the blood-brain barrier and enters cell membranes. Ascorbate (vitamin C) crosses the blood-brain barrier less easily but is actively transported into CSF by the choroid plexus and is further concentrated in neuronal cytoplasm by a second active system. Vitamin C actually exists in a relatively unreactive free radical form, which by reacting with O 2 ~- and OHradicals can prevent further propagation of chain reactions such as that illustrated in Figure 1. Vitamin E neutralizes free radicals by donating hydrogen

2 (1) AJSJU (2) /UUU o) \r\=/w (4) V = \=A=/ o o \ruu -o 2 H*+.OH H 2 O FIGURE 1. In (I), part of a polyunsaturated fatty acid (PUFA) chain is shown with three double bonds. A -OH radical "abstracts" a hydrogen atom, forming H 2 O and an alkyl free radical center on the PUFA molecule (2), which then rearranges to form (3) a "conjugated diene" (a saturated C no longer stands between C atoms with double bonds). Molecular O 2 then adds to form a peroxy radical (4), which can then abstract a hydrogen atom from a second PUFA molecule, creating another free radical as in (2) and propagating the reaction. atoms; this reaction, of course, leaves an unpaired electron on the vitamin E molecule, but the free radical thus created is harmless. Specific enzymes have also evolved to deal with free radicals. Superoxide dismutase, which exists in mitochondrial and cytoplasmic forms, catalyzes the conversion of 2 O 2 ~- molecules into H 2 O 2 and O 2. Because H 2 O 2 is a potential source of OH- radical, two additional protective enzymes, catalase and glutathione peroxidase, destroy it. The former converts H 2 O 2 molecules into O 2 and H 2 O, and the latter catalyzes the oxidation of reduced glutathione by 2 H 2 O 2. Another enzyme, glutathione reductase, then regenerates reduced glutathione. Glutathione peroxidase detoxifies lipid hydroperoxides (see below), as well as H 2 O 2. The CNS is relatively poorly endowed with superoxide dismutase, catalase, and glutathione peroxidase and is also relatively lacking in vitamin E. However, it is rich in Fe, whose important role in production of free radicals has already been discussed, and in polyunsaturated fatty acids, which are prime targets for free radical attack. 5 Free Radical Involvement in Tissue Damage in Cerebral Ischemia The bulk of experimental work on free radicals in cerebral ischemia has concentrated on damage to Schmidley Free Radicals in CNS Ischemia 1087 membrane lipids As already noted, neuronal membranes are rich in polyunsaturated fatty acids, which are particularly susceptible to free radical attack at carbons adjacent to double bonds. Figure 1 indicates how this is thought to occur. The lipid hydroperoxides shown in Figure 1 are not completely stable in vivo and, in the presence of metals or metal complexes (e.g., Fe), can further decompose to reactive radicals which will propagate the chain reactions started by the initial free radical attack. Lipid hydroperoxides also fragment to produce aldehydes which can in turn cross-link proteins, rendering them useless as receptors or enzymes. Potential consequences of damage to membrane lipids include changes in fluidity and permeability and in the orientation of proteins embedded in the bilayer of the plasma membrane and other cellular endomembranes. The shape and function of many intrinsic membrane proteins depend on specific interactions between their hydrophobic domains and membrane phospholipids. 511 The function of these phospholipid-dependent proteins would be particularly susceptible to free radical damage of the fatty acid component of membrane phospholipids. 11 It is not surprising that receptor function is also compromised by lipid peroxidation. 12 The consequences of these alterations for cellular function are potentially lethal. Because they are extremely short lived and produced only in minute quantities, free radicals are difficult to measure directly. 12 Alternative approaches to demonstrate their involvement in cerebral ischemic damage have concentrated on measuring the rate of consumption of endogenous protective molecules, such as vitamins C and E, and reduced glutathione or the formation of by-products of lipid peroxidation, such as malondialdehyde (MDA), or conjugated dienes (Figure 1). The MDA assay has several shortcomings, not the least being its lack of specificity for free radical mediated injury. 7 Lastly, in many experimental systems free radical involvement has been inferred from the protective effects of vitamins C or E, iron chelators such as deferoxamine, inhibitors of lipid peroxidation, or enzymes like superoxide dismutase and catalase. In view of the need to rely on these indirect or inferential methods and of the well-known vagaries of animal models of cerebrovascular disorders, it is not surprising that unequivocal evidence of free radical mechanisms in cerebral ischemia has been hard to come by. More recently, "trapping" techniques have been applied to the study of free radicals in cerebral ischemia. These techniques rely on the reaction of free radicals with reagents to produce longer lived products which can be detected using electron spin resonance spectroscopy or high performance liquid chromatography with electrochemical detection However, even these newer techniques may not be free of artifact, particularly in biologic systems. 15 In vitro incubation of brain slices, cell cultures, homogenates, or subcellular fractions with free radical generating systems results in various combinations of

3 1088 Stroke Vol 21, No 7, July 1990 diene conjugation, MDA production, destruction of membrane phospholipids with concomitant release of free fatty acids, particularly polyunsaturated fatty acids, and depletion of vitamin E. These changes are accompanied by histologic, ultrastructural, and neurochemical evidence of tissue damage and edema. The activity of both neuronal and blood-brain barrier endothelial Na + -K + ATPase is impaired, as are synaptosomal uptake of serotonin and GABA and mitochondrial function. These results (summarized in reference 12) establish that free radicals are clearly capable of damaging CNS tissue but leave unanswered questions about whether (and where) free radical mechanisms contribute to cerebral ischemic damage in vivo. We have not yet addressed the question of how O 2 -derived free radicals might be generated in ischemic tissues. In nearly complete ischemia insufficient O 2 is available to accept electrons passed along the mitochondrial electron transport chain, leading to eventual reduction ("electron saturation") of components of this system, such as flavin adenine dinucleotide and coenzyme Q (CoQ). In the presence of small amounts of O 2, these molecules can then auto-oxidize to produce, for example, O 2 ~* : CoQ (reduced)+o 2 ->CoQ-+O 2 ~\ The residual O 2 molecules in severely ischemic brain cannot act as electron acceptors in the "normal" fashion because oxidationreduction ("redox") potential sufficient to favor stepwise electron transfer to them cannot be generated by such low concentrations of molecular oxygen. 11 With reperfusion reactive oxygen radicals may be generated as by-products of the reactions of arachidonic acid to produce prostaglandins and leukotrienes. These enzymes would be particularly active during reperfusion because an abundance of their substrate, free arachidonic acid, would have been released from membrane phospholipids during ischemia, and with reperfusion O 2 would also be available. During reduction of the hydroperoxide group of prostaglandin G 2 (PGG 2 ) to form prostaglandin H 2 (PGH 2 ), a free radical intermediate forms which in turn is reduced by nicotinamide adenine dinucleotide (NAD) or NAD-phosphate, generating free radical forms of these compounds which could then donate an electron to O 2 to form O 2 ~\ The issue of reperfusion injury to brain or its microvessels assumes new importance in view of recently developed thrombolytic therapies for CNS ischemia. Experiments using depletion of vitamin C, reduced glutathione, or MDA production as indicators of in vivo free radical production in ischemic CNS tissues have produced contradictory results. Recent work using the more sensitive conjugated diene technique has demonstrated lipid peroxidation during recirculation following global and focal cerebral ischemia. However, the results suggested spotty, focal involvement and were more impressive for global than for focal ischemia. 910 The "trapping" techniques described earlier have been used by two groups recently to detect free radical production after 8 minutes of global ischemia and 15 minutes of reperfusion in the rat 13 and after 5-15 minutes of severe forebrain ischemia and 5-15 minutes of reperfusion in the gerbil. 14 Thus, conclusive evidence that free radicals play an important direct role in focal cerebral ischemia is still lacking at this time. The part played by the O 2 ~- generating enzyme, xanthine oxidase (XO), in cerebral ischemia has also not been defined. In ischemia of other viscera including gut and heart, xanthine dehydrogenase (XDH), an enzyme which ordinarily cannot produce free radical species, is converted to XO, probably by Ca 2+ -activated proteolysis. 3 With reperfusion XO then catalyzes the conversion of hypoxanthine to urate, forming O 2 ~* as a byproduct. O 2 ~* can, of course, be scavenged by superoxide dismutase but with the creation of a molecule of H 2 O 2, which, as we have seen, can react with further O 2 ~- molecules in an iron-catalyzed reaction to form OH-. Substrate for XO is present in abundance because it accumulates as ATP is depleted. Brain content of XDH/XO is low, but the enzyme is present in the microvessels of the blood-brain barrier. 16 Cerebral Vasculature as Target for Free Radicals Up to this point we have considered only the brain parenchyma as a target for free radical damage. The experiments of Kontos and colleagues, however, strongly suggest that free radical mechanisms are important in mediating the response of small cerebral arterioles to acute hypertension and perhaps in the changes of hypertensive encephalopathy. These studies began with the observation that cyclooxygenase inhibitors eliminated and direct application of arachidonic acid reproduced the cerebral arteriolar responses to acute hypertension, including dilatation, diminished responsiveness to CO 2, morphologic changes in endothelium and smooth muscle, and increased permeability. These responses could be elicited not only by topical arachidonic acid but also by PGG? and 15 HPETE (15 hydroperoxyeicosatetraenoic acid), substrates for hydroperoxidase enzymes, whereas the stable PGs and 15 HPETE had no effect. The hypothesis that O 2 ~- radical produced during the PG hydroperoxidase reactions was responsible was further strengthened by the demonstration that superoxide dismutase and catalase blunted or eliminated the arterial responses induced by topical arachidonic acid, PGG 2, and 15 HPETE. It was proposed that in acute hypertension release of arachidonic acid from phospholipids is the initial event, stimulated by activation of phospholipase A 2. The arachidonic acid is then metabolized by cyclooxygenase to PGG 2 and thence to PGH 2 by PG hydroperoxidase, producing O 2 ~\ The charged O 2 ~* molecules probably escape via an anion channel in the cell membrane to enter the CSF and brain extracellular fluid. More recently, the same group has demonstrated O 2 ~' production in vascular smooth muscle and endothelium during reperfusion following complete ischemia. 19

4 The effects of O 2 on the microvessels of the CNS have been studied in capillaries of the frog, which, in addition to being somewhat larger than those of mammals, are present on the pial surface as well as within the neural parenchyma. These features allow penetration of intact microvessels with microelectrodes and determination of the electrical resistance of the endothelial layer. 20 Topical application of xanthine/xo to the pial surface of the frog brain causes a reversible decrease in electrical resistance of the endothelium, presumably reflecting increased permeability of tight junctions. The changes induced by xanthine/xo were inhibitable by allopurinol and superoxide dismutase+ catalase but not by corticosteroids, which act by inhibiting phospholipase A 2, nor by cyclooxygenase or lipoxygenase inhibitors, suggesting that they were evoked by O 2 ~ g radicals generated by XO, rather than via arachidonic acid pathways. 21 The relevance of these free radical mechanisms to human hypertensive encephalopathy, loss of autoregulation in ischemic tissue, ischemic brain edema, or vasospasm following subarachnoid hemorrhage (SAH) is unsettled but of potential interest. Although the arteriolar change evoked by free radicals in this experimental system is dilation, it is possible that arteriolar smooth muscle damage following SAH might be enough to produce a necrotizing arteriopathy. Post-SAH CSF would contain abundant iron to catalyze production of OH- radicals. Recent Developments in Therapy The 21-aminosteroids are a recently synthesized class of molecules which, because of the 21-amino substitution, lack glucocorticoid or other "classic" steroid hormone activity. They cross the blood-brain barrier after systemic administration and are powerful inhibitors of lipid peroxidation of CNS tissues in vitro. Intraperitoneal injection of gerbils with these agents, U74006F and U74500A, before and after unilateral carotid occlusion protects hippocampal CA! neurons and certain neocortical areas against ischemic damage and, at high doses, improves survival This protective effect was not seen in a 15-minute bilateral carotid occlusion model in the gerbil. 23 In the rat MCA occlusion model, U74006F given at 10 minutes and again at 3 hours after occlusion reduced ischemic brain edema. 24 In the cat posttreatment with U74006F attenuated postischemic hypoperfusion following 5 minutes of global ischemia, which suggested a protective effect on the microcirculation. However, in this system animals given U74006F also had much higher blood pressures in the postischemic period, an effect not noted in other models, which may have increased cerebral perfusion pressure. 25 U74006F is capable of scavenging and thereby neutralizing the potentially deleterious effects of O 2 ~- and lipid hydroperoxides. It also blocks the release of arachidonic acid from membranes, which may also indirectly contribute to generation of free radicals (see above). 21 U74500A has iron-binding properties which may explain its greater Schmidley Free Radicals in CNS Ischemia 1089 efficacy in vitro against lipid peroxidation. 23 It remains to be shown that any 21-aminosteroid actually reduces free radical generation or lipid peroxidation in vivo. The exact locus (microcirculation versus neuron) and mechanism of action of these exciting new molecules still must be elucidated. Although superoxide dismutase has been used to modify ischemia-reperfusion injury of the lung, gut, and myocardium, several obstacles seem to block its use in cerebral ischemia. These include a negative charge and high molecular weight, which mean exclusion by the blood-brain barrier, as well as a short plasma half-life and the inability of neurons and astrocytes to internalize the enzyme. The bloodbrain barrier problem can be circumvented by using liposome-entrapped enzyme. This approach has been used in focal models of cerebral ischemia with reduction of edema and infarct size. 26 Another potential delivery system, consisting of superoxide dismutase or catalase conjugated to polyethylene glycol, has also been effective in reducing the size of focal ischemic brain lesions This review ends on a somewhat paradoxic note, reporting exciting new potential therapies which seem to act by scavenging free radicals or blocking lipid peroxidation in CNS. Yet unequivocal evidence supporting production of free radicals in cerebral ischemia has only recently emerged, and proof of their importance vis a vis other putative mechanisms in focal CNS ischemia, such as Ca 2+ influx, acidosis, and excitatory neurotransmitters, is still lacking. 29 Perhaps one way to reconcile these seemingly disparate results is the concept that free radical mechanisms damage the microvasculature, primarily during reperfusion, causing further parenchymal damage as a result of increased permeability or platelet aggregation. 19 Since the microvasculature is quantitatively a very minor fraction of the brain mass, even severe damage might not be evident on sensitive assays for the products of lipid peroxidation, which would explain the largely negative results reported above. 30 Acknowledgments The author wishes to thank Drs. P.H. Chan and R.A. Fishman for nurturing his interest in this subject, Dr. M.D. Winkelman for a critical review of the manuscript, and Barbara First for tireless and technically superb word processing. References 1. Cross CE, Halliwell B, Borish ET, Pryor WA, Ames BN, Saul RL, McCord JM, Harman D: Oxygen radicals and human disease. Ann Intern Med 1987;107: Southorn PA: Free radicals in medicine: II. Involvement in human disease. Mayo CM Proc 1988;63: Cord JM: Oxygen-derived free radicals in postischemic tissue injury. NEnglJMed 1985;312: Freeman BA, Crapo JD: Biology of disease: Free radicals and tissue injury. Lab Invest 1982;47: Halliwell B, Gutteridge JMC: Oxygen radicals and the nervous system. Trends Neurosci 1985;8: Southorn PA, Powis G: Free radicals in medicine: I. Chemical nature and biologic reactions. Mayo Clin Proc 1988;63:

5 1090 Stroke Vol 21, No 7, July 1990 Downloaded from by on November 22, Halliwell B, Gutteridge JMC: Oxygen free radicals and iron in relation to biology and medicine: Some problems and concepts. Arch Biochem Biophys 1986;246: Slater TF: Free-radical mechanisms in tissue injury. Biochem J 1984;222: Ginsberg MD, Watson BD, Busto R, Yoshida S, Prado R, Nakayama H, Ikeda M, Dietrich WD, Globus MY-T: Peroxidative damage to cell membranes following cerebral ischemia: A cause of ischemic brain injury. Neurochem Pathol 1988; 9: Watson B, Ginsberg MD: Mechanisms of lipid peroxidation potentiated by ischemia in brain, in Halliwell B (ed): Oxygen Radicals and Tissue Injury. Bethesda, Md, Federation of American Societies for Experimental Biology, 1988, pp Demopoulos HB, Flamm ES, Seligman ML, Mitamura JA, Ransohoff J: Membrane perturbations in central nervous system injury: Theoretical basis for free radical damage and a review of the experimental data, in Popp AJ, et al (eds): Neural Trauma. New York, Raven Press, Publishers, 1979, pp Chan PH: The role of oxygen radicals in brain injury and edema, in Chow CK (ed): Cellular Antioxidant Defense Mechanisms, Volume III. Boca Raton, Fla, CRC Press, Inc, 1988, pp Kirsch JR, Phelan AM, Lange DG, Traystman RJ: Reperfusion-induced free radical formation following global ischemia (abstract). Pediatr Res 1987;21:202A 14. Cao W, Carney JM, Duchon A, Floyd RA, Chenon M: Oxygen free radicals involvement in ischemia and reperfusion injury to brain. Neurosci Lett 1988;88: Zweier JL, Flaherty JT, Weisfeldt ML: Measurement of free radical generation in the post-ischemic heart, in Cerutti PA, Fridorich I, McCord JM (eds): Oxy-Radicals in Molecular Biology and Pathology. New York, Alan R. Liss, 1988, pp Betz AL: Identification of hypoxanthine transport and xanthine oxidase activity in brain capillaries. J Neurochem 1985; 44: Kontos HA: Oxygen radicals in cerebral vascular injury. Circ Res 1985;57: Kukreja RC, Kontos HA, Hess ML, Ellis EF: PGH synthase and lipoxygenase generate superoxide in the presence of NADH or NADPH. Circ Res 1986;59: Kontos HA: Oxygen radicals in cerebral ischemia, in Ginsberg MD, Dietrich WD (eds): Cerebrovascular Diseases, 16th Princeton Conference. New York, Raven Press, Publishers, 1989, pp Crone C, Olesen SP: Electrical resistance of brain microvascular endothelium. Brain Res 1982;241: Olesen SP: Free oxygen radicals decrease electrical resistance of microvascular endothelium in brain. Ada Physiol Scand 1987;129: Hall ED, Pazara KE, Braughler JM: 21-aminosteroid lipid peroxidation inhibitor U74006F protects against cerebral ischemia in gerbils. Stroke 1988;19: Hall ED, Pazara KE: Effects of novel 21-aminosteroid antioxidants on postischemic neuronal degeneration, in Ginsberg MD, Dietrich WD (eds): Cerebrovascular Diseases, 16th Princeton Conference. New York, Raven Press, Publishers, 1989, pp Young W, Wojak JC, DeCrescito V: 21-aminosteroid reduces ion shifts and edema in the rat middle cerebral artery occlusion model of regional ischemia. Stroke 1988;19: Hall ED, Yonkers PA: Attenuation of postischemic cerebral hypoperfusion by the 21-aminosteroid U74006F. Stroke 1988; 19: Chan PH, Chen S, Imaizumi S, Chu L, Chan T: New insights into the role of oxygen radicals in cerebral ischemia, in Bazan NG, Braquet P, Ginsberg M (eds): Neurochemical Correlates of Cerebral Ischemia. New York, Plenum Publishing Corp (in press) 27. Liu TH, Beckman JS, Freeman BA, Hogan EL, Hsu CY: Polyethylene glycol-conjugated superoxide dismutase and catalase reduce ischemic brain injury. Am J Physiol 1989; 256:H589-H Beckman JS, Liu TH, Hogan EL, Lindsay SL, Freeman BA, Hsu CY: Evidence for a role of oxygen radicals in cerebral ischemic injury, in Ginsberg MD, Dietrich WD (eds): Cerebrovascular Diseases, 16th Princeton Conference. New York, Raven Press, Publishers, 1989, pp Siesjo BK: Mechanisms of ischemic brain damage. Crit Care Med 1988;16: Watson BD: What is the evidence for oxygen radical-mediated reperfusion injury in stroke? in Ginsberg MD, Dietrich WD (eds): Cerebrovascular Diseases, 16th Princeton Conference. New York, Raven Press, Publishers, 1989, pp KEY WORDS cerebral ischemia free radicals