I. Introduction

Top Previous Next References

Stroke is a sudden loss of brain function resulting from interference with the blood supply to the central nervous system (CNS¹). Acute stroke can be classified either as ischemic (80% of stroke cases), which can be further classified to extra-cranial embolism and intracranial thrombosis, or a hemorrhagic stroke (20% of stroke cases), which can be further classified to intracerebral hemorrhage and subarachnoid hemorrhage (SAH; Fig. 1).

View larger version (9K): [in this window] [in a new window]   Fig. 1.   Classification of acute stroke.

Stroke is the third most common cause of death in Europe and North America, and is a major cause of morbidity particularly in the middle-aged and elderly population (Bronner et al., 1995; De Freitas and Bogousslavsky, 2001). CNS damage occurs in stroke as a result of hypoxia. In cerebral ischemia there is an ischemic gradient that can be divided into the core, which is the central ischemic zone, and the penumbra, which is located in more peripheral zones. In the penumbra, functional alterations occur in the neurons and glial cells. Neurons are most vulnerable to hypoxia due to their dependence on the oxidative metabolism of glucose for energy. The principal pathophysiological processes in acute CNS injury, such as stroke, mechanical trauma, or subarachnoid hemorrhage, are extremely complex and involve pathological permeability of blood brain barrier (BBB, in part of the CNS injuries), energy failure, loss of cell ion homeostasis, acidosis, increased intracellular calcium, excitotoxicity, and free radical-mediated toxicity. This can lead to ischemic necrosis, which occurs in the severely ischemic regions and is associated with loss of calcium and glutamate homeostasis. It can also lead to apoptosis, which is more likely to occur in the moderately ischemic regions, evolves more slowly, and depends on the activation of a sequence of genes (Pulsinelli, 1992; Gennarelli, 1997; Dirnagl et al., 1999; Fig. 2)

View larger version (20K): [in this window] [in a new window]   Fig. 2.   Cellular mechanisms that may be involved in acute ischemia and CNS injury.

A free radical is any chemical compound that contains one or more unpaired electrons in its outer orbits. Unpaired electrons alter the chemical reactivity of an atom or molecule, usually making it more reactive than the corresponding nonradical, because they act as an electron acceptor and essentially "steal" electrons from other molecules. This electron loss is called oxidation and free radicals are referred to as oxidizing agents (Halliwell and Gutteridge, 1989). Humans are constantly exposed to free radicals created by external sources from the environment (e.g., radon and cosmic radiation) or man-made and by internal cellular metabolisms. The most commonly occurring cellular free radical is superoxide radical (O), which is produced when an oxygen molecule gains one electron from another substance. Excess amount of O leads to tissue damage by promoting hydroxyl radical (OH^·) formation through hydrogen peroxide (H₂O₂) (the iron-mediated Haber-Weiss reaction; Jenner and Olnaw, 1996; Simonian and Coyle, 1996; Fig. 3A). Alternatively, O may lead to OH^· formation through an interaction with endogenously formed nitric oxide (NO^·) by nitric oxide synthetase, an enzyme that is concentrated in certain neurons and activated by Ca²⁺. The interaction of NO^· with O leads to the formation of peroxynitrite (ONOO), which can generate nitrosyl radical (ONOOH), which decomposes to form OH^· (Fig. 3B). Free radicals and related molecules are often classified together as reactive oxygen species (ROS) to signify their ability to lead to oxidative changes within the cell (Simonian and Coyle, 1996). These radicals can cause cellular damage to cardinal cellular components such as lipids. Polyunsaturated fatty acids are particularly vulnerable to free radical attack, because the double bonds within membranes allow easy removal of hydrogen ions by ROS such as OH^· (Halliwell and Gutteridge, 1989). Free radicals can also damage proteins and nucleic acids (e.g., DNA), leading to subsequent cell death by mode of necrosis or apoptosis. Cells normally have a number of mechanisms acting to defend against damage induced by free radicals (Evans, 1993; Simonian and Coyle, 1996). Problems occur when production of ROS exceeds their elimination by the antioxidant protective systems or when the latter are damaged. This imbalance between cellular production of ROS and the inability of cells to defend against them is called oxidative stress (OS) (Ebadi et al., 1996; Jenner and Olnaw, 1996; Simonian and Coyle, 1996). OS is involved in acute and chronic CNS injury and is a major factor in the pathogenesis of neuronal damage (Facchinetti et al., 1998).

View larger version (25K): [in this window] [in a new window]   Fig. 3.   Reactive oxygen species reactions that may lead to OH· production. NADP, nicotinamide adenine dinucleotide phosphate; NADPH, reduced form of NADP; NOS, nitric oxide synthetase.

1. Oxidative Stress-Mediated Brain Damage. Some of the pathological processes in acute CNS injury involve the generation of oxygen free radicals either as a cause or a result of disease progression (Love, 1999; Lewen et al., 2000). Free radicals are generated by the constant use of oxygen in the mitochondria to supply the energy needs of the brain. Some enzymes expressed in the brain including monoamine oxidase, tyrosine hydroxylase, and L-amino acid oxidase produce H₂O₂ as a normal byproduct of their activity. The activity of other neuronal enzymes yields oxidants such as the Ca²⁺-dependent activation of phospholipase A₂. That may lead to arachidonic acid release, producing O through its subsequent metabolism by lipoxygenases and cyclo-oxygenases to form eiconasoids. Auto-oxidation of endogenous substances, e.g., ascorbic acid and catecholamines, may yield high levels of H₂O₂ (Coyle and Puttfarcken, 1993). Therefore, ROS have been the focus of interest as possible candidates for the elicitation of various pathological responses in the pathogenesis of acute CNS injury and as a therapeutic target (Bromont et al., 1989; Hall, 1989; Oliver et al., 1990). It is well known that glial cells are more resistant to OS than neurons, probably due to transcriptional up-regulation of glutathione synthesis (Rice and Russo-Menna, 1998; Iwata-Ichikawa et al., 1999).

2. Excitotoxicity Insults. One of the first pathophysiological events leading to neuronal damage in acute CNS injury involves glutamate accumulation in the extracellular space. Glutamate is the major excitatory amino acid among the excitatory amino acids (EAA) in the brain, acting mainly through activation of its ionotropic receptors. These receptors can be distinguished by their pharmacological and electrophysiological properties: the -amino-3-hydroxy-5-methyl-4-isoxasole-proprionic acid, kainic acid, and the N-methyl-D-aspartate (NMDA) receptors. Activation of these receptors leads to depolarization and neuronal excitation. However, if for any reason receptor activation becomes excessive or prolonged, the target neurons become damaged and eventually die. In the ischemic brain, extracellular glutamate is elevated rapidly after the onset of ischemia and declines after reperfusion. The mechanisms that are responsible for the elevation of extracellular glutamate include enhanced efflux of glutamate and the reduction of glutamate uptake. This process seems to involve sustained elevations of intracellular calcium levels through glutamate transporters operating in the reverse mode, and owing to imbalance of sodium ions across plasma membranes. Moreover, the fact that the brain can neither synthesize nor store energy reserves, means that any interruption in cerebral blood flow may lead to rapidly and irrevocably energy failure and dramatic fall in intracellular levels of ATP. The consequences will be an increase in the concentrations of extracellular glutamate and neuronal sensitization to excitotoxic cell death. It is well established that high levels of glutamate in the extracellular space appear rapidly after the onset of ischemia. Nevertheless, a direct linkage between the enhanced release of glutamate and the neuronal injury has not been fully established (Coyle and Puttfarcken, 1993; Bondy, 1995; Doble, 1999). Pharmacological studies in rodents and recent clinical studies in humans have shown that the extra-neuronal concentration of glutamate rose to toxic levels under ischemic (Benveniste et al., 1984; Hagberg et al., 1985; Siesjo, 1992a,b; Bullock et al., 1995; Davalos et al., 1997) and traumatic (Faden et al., 1989) conditions. In addition, NMDA antagonists that were added to neuronal cultures, rescued cells treated with glutamate receptor agonists. Neuroprotection can be achieved by blocking the presynaptic release of glutamate and/or by blocking the excitation of postsynaptic neurons occurring after an ischemic episode. In this regard, the voltage-sensitive calcium channels and glutamate receptors may be suitable targets for therapy (Coyle and Puttfarcken, 1993; Nishizawa, 2001).

3. Oxidative Stress and Excitotoxicity. It is well known that EAA and neurotransmitters, whose metabolism produces ROS, are unique in the brain as sources of OS (Coyle and Puttfarcken, 1993). It has been proposed that during CNS, ROS (mainly O and NO^·) and EAA (mainly glutamate) may cooperate in the pathogenesis of neuronal damage, involving loss of cellular calcium homeostasis (White et al., 1984; Bose et al., 1992; Siesjo, 1993).

4. Oxidative Stress-Induced Gene Expression. A large number of gene products appear after an ischemic insult, making it difficult to decipher which genes are really involved in the mechanism of tissue injury. ROS were shown to influence gene expression and to play a role in the events that lead to neuronal death. In global cerebral ischemia, the oxidative responsive transcription factor, nuclear factor-B (NF-B), is persistently activated. Overall, persistent NF-B activation enhances ischemic neuronal death (Schneider et al., 1999), but its effects differ between cell types. Activation of NF-B in neurons induces production of anti-apoptotic gene products and proteins involved in modulating synaptic plasticity and increases their survival after stroke. Activation of NF-B in glial cells (astrocytes and microglia) results in the production of proinflammatory cytokines and potentially neurotoxic ROS and excitotoxins, thus, promoting ischemic neuronal degeneration (Mattson and Camandola, 2001). The NF-B translocation into the nucleus and binding to the NF-B site can activate many inducible genes, including but not only cyclooxygenase-2, inducible nitric oxide, metalloproteinases, intercellular adhesion molecules, and cytokines. The expression of these genes may lead to formation of ROS and BBB breakdown, which may lead to apoptosis or necrosis or both (Chan, 2001). In addition to NF-B, many transcription factors such as AP-1, HIF-1, SP-1, and EIK-1 are known to be redox-sensitive proteins (Sen, 1998), and their regulation of gene expression by OS in cerebral ischemia has yet to be determined (Sharp et al., 2000).

The BBB is a major barricade that separates the brain microenvironment from the blood within the cerebrovascular tree to allow complex neural signaling without external interference. According to ultrastructural studies, endothelial cells in the brain differ fundamentally in two ways from those in peripheral tissues. First, they have very few endocytotic vesicles, limiting the amount of trans-cellular flux. Second, they are coupled by tight junctions or a zipper-like structure that seal the intercellular cleft and restrict par-cellular flux (Reese and Karnovsky, 1967). Normally, the tight junctions of the BBB permit the diffusion of only very small amounts of water-soluble compounds (par-cellular aqueous pathway), whereas the large surface area of the lipid membranes of the endothelium offers an effective diffusive route for lipid-soluble agents (trans-cellular lipophilic pathway; Rowland et al., 1992). Pathological permeability of BBB may occur after closed head injury (CHI) and SAH, enabling easier drug penetration to the brain. In other acute CNS injuries such as cerebral ischemia, the BBB is intact, at least in part, leading to reduction in drug penetration into the brain. This problem has prompted researchers to develop methods to induce transient opening of the tight junctions of the brain endothelial cells, such as osmotic opening with mannitol or arabinose (Gumerlock and Neuwelt, 1992).

	II. Antioxidants in the Treatment of Acute Central Nervous System Injury

Top Previous Next References

Antioxidants are exogenous (natural or synthetic) or endogenous compounds acting in several ways including removal of O₂, scavenging reactive oxygen species or their precursors, inhibiting ROS formation and binding metal ions needed for catalysis of ROS generation. The natural antioxidant system can be classified into two major groups: enzymes and low molecular weight antioxidants (LMWA). The enzymes include SOD, catalase, peroxidase, and some supporting enzymes. The LMWA group of molecules can be further classified into directly acting antioxidants (e.g., scavengers and chain breaking antioxidants) and indirectly acting antioxidants (e.g., chelating agents). The former are extremely important in defense against OS. This subgroup currently contains several hundred compounds. Most of them, including ascorbic and lipoic acids, polyphenols, and carotenoids, are derived from dietary sources (Shohami et al., 1997). The cell itself synthesizes a minority of these molecules, such as glutathione and NADPH. The distribution of protective antioxidants in the body has some interesting features. For instance, there is a relatively high concentration of the water-soluble antioxidant vitamin C in the brain. However, vitamin E concentrations in CNS are not remarkably different from those in other organs. The concentrations of antioxidants also vary within the different regions of the brain itself. For instance, the lowest concentration of vitamin E is found in the cerebellum (Vatassery, 1992). It was also shown that enzymatic antioxidants, such as catalase, are in lower concentrations in the brain than in other tissues.

As with other neuroprotectants, to achieve high efficacy, antioxidants must penetrate through the BBB, and be given as early as possible and within the "neuroprotective window" (the time interval where they significantly reduce or prevent cerebral damage). The therapeutic window for successful attenuation of an infarct volume was shown to be 3 to 4 h in rats (Kaplan et al., 1991; Memezawa et al., 1992) and cats (Heiss and Rosner, 1983) and 6 to 8 h in nonhumans primates (Jones et al., 1981). Current early-phase trials of neuroprotectants in stroke (e.g., NMDA antagonists) adhered to the 4- to 6-h time frame within which tissue rescue may be possible (Ginsberg, 1994; Pulsinelli, 1995). This is supported by statistical analysis of relevant animal studies suggesting that irreversible focal injury begins within a few minutes and is complete within 6 h (Zivin, 1998).

1. Vitamins. a. In Prevention. An important finding of epidemiological studies on stroke is the lower risk of cerebral ischemic events among individuals with frequent consumption of fruit and vegetables (Acheson and Williams, 1983; Vollset and Bjelke, 1983; Joshipura et al., 1999). The specific nutrients responsible for this effect remain elusive, but antioxidant vitamins, such as vitamin E, -carotene, and vitamin C, which are free radical scavengers, may be major contributing factors to this phenomenon.

2. Coenzyme Q₁₀. Coenzyme Q₁₀ (ubiquinone) is a mobile and lipid-soluble compound within the hydrophobic core of the phospholipid bilayer of the inner membrane of the mitochondria. It is an essential cofactor in the electron transport chain, where it accepts electrons from complexes 1 and 2 (Beyer, 1992; Ernster and Dallner, 1995; Do et al., 1996). Coenzyme Q₁₀ also serves as an important antioxidant in lipid membranes (Noack et al., 1994; Forsmark et al., 1997) either directly or by regenerating vitamin E. Its levels are known to decrease with age in both human and rat tissues (Beyer et al., 1985; Kalen et al., 1989; Battino et al., 1995). This decrease may be caused by reduced synthesis or age-dependent increases in lipid peroxidation (Forsmark et al., 1997).

3. Melatonin. Melatonin (N-acetyl-5-methoxytryptamine) is an indoleamide secreted by the pineal gland, which has structural similarities to serotonin. Melatonin is known as a biological modulator of many physiological mechanisms (e.g., circadian rhythms and sleep). It is highly lipophilic and, when administered exogenously, can readily cross the BBB and gain access to neurons and glial cells. There is experimental evidence that melatonin influences aging and age-related processes and disease states (Beyer et al., 1998). These roles are apparently related to its potency as a free radical scavenger (Beyer et al., 1998).

4. -Lipoic Acid. -Lipoate is a LMWA absorbed from the diet, which crosses the BBB (Packer, 1992; Packer et al., 1997). It is intracellularly reduced to dihydrolipoate, which is exported to the extracellular medium. Both -lipoate and especially dihydrolipoate are potent antioxidants and reduce lipid peroxidation. Hence, protection is potentially afforded to both intracellular and extracellular environments.

5. Ebselen. Glutathione peroxidase, in both selenium-dependent and -independent forms, is one of the major enzymes responsible for the degredation of hydrogen peroxide and organic peroxides in the brain. The seleno-organic compound ebselen has antioxidant activity through a glutathione peroxidase-like action. This data has led to extended research of this molecule (Muller et al., 1984; Wendel et al., 1984).

6. Human Superoxide Dismutase/Superoxide Dismutase-Like Molecules. SOD converts superoxide to hydrogen peroxide (H₂O₂) and represents the first line of defense against oxygen toxicity. Three forms have been described in man: The first isoform, containing copper and zinc at its active site (Cu/Zn SOD-1), is found in the cytoplasm of cells. Another isoform, containing manganese at its active site, is located in the mitochondria (Mn SOD-2). The third isoform is present in extracellular fluids such as plasma (Cu/Zn SOD-3). It was found that the traces of copper, zinc, and manganese metals are essential for maintaining the antioxidant activity of SOD (Halliwell, 1994).

7. Spin-Trap Scavenging Agents. Spin-trap scavenging agents are molecules (usually with a nitrone moiety) that have been used in electron paramagnetic studies for trapping highly reactive, unstable radicals. These compounds have been shown to protect experimental animals from pathology, mainly associated with ischemia-reperfusion injury, physical trauma, and aging (Carney et al., 1991; Hensley et al., 1997). Phenyl--tert-butyl nitrone (PBN) is a synthetic antioxidant capable of scavenging oxygen- and carbon-based free radicals (Kotake, 1999).

8. N-Acetylcysteine. NAC is a thiol-containing compound used in clinical practice since the mid-1950s. NAC has been demonstrated to effectively reduce free radical species and other oxidants, especially OH and H₂O₂ (Moldeus et al., 1986). NAC is not synthesized endogenously and cannot cross the BBB after exogenous administration. This fact limits the efficacy of NAC in vivo.

9. Glutathione. GSH is a ubiquitous tri-peptide formed from three amino acidsglutamate, glycine, and cysteineand synthesized by two ATP-dependent enzymatic reactions (Richman and Meister, 1975; Meister and Anderson, 1983). It can also be generated from metabolism of NAC. GSH has major intracellular antioxidant activity, mainly due to the thiol group within the molecule. It plays a critical role in detoxification of peroxides and electrophilic toxins as a substrate for GSH peroxidase and GSH transferase (Larsson et al., 1983; Meister and Anderson, 1983).

10. Metal Ion Chelators. Free metal ions are associated with the pathology of various neurodegenerative diseases (e.g., copper in Wilson's disease and iron in the substantia nigra in Parkinson's disease). Therefore, proteins that are involved in the binding of metal ions were suggested to act as antioxidants. These may include transferrin (binds iron), ceruloplasmin (binds copper), and hemopexin (binds heme, a catalyst in oxidative reactions). Desferral, a potent chelator of redox-active metals, was shown to facilitate the clinical recovery of traumatized rats in a model of CHI (Zhang et al., 1998). Deferoxamine, an iron catalyst in the generation of free radicals and lipid peroxides, given prior or soon after the ischemic episode improved survival and physiological functions in rats (Palmer et al., 1994), dogs (Hurn et al., 1995), and mice (Sarco et al., 2000). However, Fleischer et al. (1987) did not find any benefit of deferoxamine in complete cerebral ischemia in dogs.

11. Uric Acid. Uric acid is a waste product of the living cell, which is produced by xanthine oxidase. It is widely distributed in relatively high concentrations throughout the body. Urate contributes up to 60% of the total plasma antioxidant activity in healthy subjects (Wayner et al., 1987, Benzie, 1996). It acts as an antioxidant by interacting with 10 to 15% of the hydroxyl radicals produced daily and by efficiently scavenging both peroxyl radicals and singlet oxygen (Ames et al., 1981). It also binds iron (Davies et al., 1986) and acts indirectly by stabilizing plasma ascorbate (Sevanian et al., 1991). In contrast, Benzie and Strain (1996) hypothesized that urate at high concentrations acts as a pro-oxidant and suggested that hyperuricemia is a risk factor for oxidative stress-associated disorders.

12. Creatine. Creatine (N-[aminoiminomethyl]-N-methyl glycine) is a tri-peptide endogenously produced from glycine, methionine, and arginine in the liver, kidney, and pancreas (McArdle et al., 1999). Creatine can be found in the muscle, but also in brain tissue (Mujika and Padilla, 1997). Recent experimental findings have demonstrated that creatine provides significant neuroprotection against ischemic and oxidative insults (Holtzman et al., 1998; Balestrino et al., 1999). Sullivan et al. (2000) showed that chronic administration of creatine ameliorated the extent of cortical damage by as much as 36% in mice and 50% in rats after experimental traumatic brain injury. The protection seemed to be related to creatine-induced maintenance of mitochondrial bioenergetics. Mitochondrial membrane potential was significantly increased, intramitochondrial levels of ROS and calcium were significantly decreased, and ATP levels were maintained. This new agent should be intensively investigated before clinical studies for acute CNS injury are performed.

13. Lazaroids. Lazaroids are 21-aminosteroids derived from glucocorticosteroids, but they lack glucocorticoid and mineralocorticoid activities. They scavenge lipid peroxyl radicals and inhibit iron-dependent lipid peroxidation (Hall, 1995). Tirilazad mesylate (U-74006F), one of the lazaroid series, is a lipophilic compound with a high affinity for vascular endothelium (Hall et al., 1994). It was shown to protect the BBB against traumatic or SAH-induced permeability. The penetration of tirilazad to brain parenchyma is enhanced after acute CNS injury and disruption of the BBB (Hall et al., 1994).

14. Nicaraven. Two recent studies using nicaraven (also called AVS (±)-N,N'-propylenedinicotinamide), a hydroxyl radical scavenger, confirmed its antivasospastic and brain-protective activities, accompanied by improved cerebral blood flow and glucose use in a rat model of SAH (Germano et al., 1998; Yamamoto et al., 2000).

15. Other Antioxidants. i. 2,4-Diamino-Pyrrolo[2,3-D] Pyrimidines. In vivo models of oxidative injury in mice (Hall et al., 1997) and ischemia models in rats (Schmid-Elasesser et al., 1997) have recently shown some efficacy of the novel 2,4-diamino-pyrrolo [2,3-D] pyrimidines. These molecules, administered orally, were identified as having a much greater BBB penetration capacity (Hall et al., 1997) and high-lipophilic antioxidant activity with protective effects (Bundy et al., 1995).

	III. Conclusion and Future Strategies

Top Previous Next References

An increasing amount of evidence suggests that OS is important in either the primary or the secondary pathophysiological mechanisms underlying acute CNS injury. In addition, reduction in the endogenous antioxidant defense system due to environmental and genetic factors may contribute to OS evolution. Therefore, the discovery and development of potent antioxidant agents has been one of the most interesting and promising approaches in the search for treatment of CNS injury. Antioxidants of varying chemical structures have been investigated as therapeutic agents in the treatment of acute CNS injury (Table 1). Although some of the antioxidants showed efficiency in animal models, most of them did not show beneficial effect in clinical trials performed to date (Table 2). To achieve efficacy, the antioxidant must be given during the "time window" available between the vascular event and irreversible neuronal loss. They also should fit to the precise OS physiology, e.g., the type of ROS involved, the place of generation, and the severity of the damage. Moreover, antioxidants must penetrate the BBB to attain a critical therapeutic level within the CNS. Thus, pharmacotherapy for closed head injury and SAH is less problematic than for other acute CNS injuries, because there is obvious disruption of the BBB, enabling easier drug penetration to the brain. Potential reasons for antioxidant failure to achieve neuroprotection in clinical trials include narrow "time window", suboptimal drug dose, inappropriate drug levels at the target CNS site, and discrepancy in drug mechanism and pathophysiological processes (Table 3). Antioxidants may have differential effects in protecting nucleic acids, proteins, and lipids from free radical damage and some compounds may be preferentially localized within specific subcellular organelles. Thus, antioxidant cocktails or antioxidants combined with other drugs such as calcium antagonists, glutamate antagonists, or anti-apoptotic agents, may have more successful synergistic effects. Better understanding of the underlying pathologic al mechanisms of acute CNS injury and improvement of the molecular design of antioxidants will open a full spectrum of possibilities for treatment of various types of injuries.