OUP user menu

New concepts in reactive oxygen species and cardiovascular reperfusion physiology

Lance B Becker
DOI: http://dx.doi.org/10.1016/j.cardiores.2003.10.025 461-470 First published online: 15 February 2004


Increasingly complex behavior of free radicals and reactive oxygen species (ROS) are noted within biological systems. Classically free radicals and ROS were considered injurious, however current mechanisms describe both protective and deleterious effects. A burst of ROS has been well described with the first moments of reperfusion and is associated with injury. However ROS can also be protective as signal preconditioning protection and induce stress responses that lead to survival. ROS generation is appreciated to occur during ischemia despite the low oxygen tension, from a likely mitochondria source, and ROS-induced ROS release may amplify its signal. The burst of ROS seen during reperfusion may originate from a different cellular source than during ischemia and is not yet fully identified. ROS and cellular redox conditions regulate a large number of vital pathways (energy metabolism, survival/stress responses, apoptosis, inflammatory response, oxygen sensing, etc). While cellular systems may demonstrate reperfusion injury, whole organ and animal models continue to report contradictory results on reperfusion injury and the role of antioxidants as a therapy. Collectively, these data may offer insight into why clinical trials of antioxidants have had such mixed and mostly negative results. Future antioxidant therapies are likely to be effective but they must become: more specific for site of action, not have deleterious effects on other signaling pathways, be targeted to a specific reactive oxygen species or cellular compartment, and be “time sensitive” so they deliver the correct therapy at precisely the correct time in ischemia and reperfusion.

  • Reactive oxygen species
  • ROS
  • Oxidant stress
  • Cell signaling
  • Reperfusion injury
  • Free radicals
  • Cardioprotection

1 New concepts in reactive oxygen species and cardiovascular reperfusion physiology

During the past decade we have extended our insights into the complex effects of free radicals and reactive oxygen species (ROS) within biological systems. Free radicals are a relative new concept in biology first described only 50 years ago, at that time they were considered uniformly injurious, the source of mutagenesis, aging and cancer [1]. In later years the description of superoxide dismutase generated a new wave of excitement as many new studies advanced our understanding of basic mechanisms of ROS metabolism and the biological effects of free radicals [2]. Several important concepts have developed which will be reviewed in this paper including: (1) Mechanisms by which ROS can be both protective and deleterious. (2) Possible intracellular sources and positive feedback for ROS generation during ischemia and reperfusion. (3) Central role of ROS signaling pathways that act on additional metabolic pathways and cellular networks. (4) Consideration of cellular reperfusion injury versus the contradictory data for whole organ and animal reperfusion injury. (5) Implications for effective clinical therapies.

2 Oxygen: source of life but also source of ROS

Aerobic organisms utilize molecular oxygen as a terminal electron acceptor to enable metabolism of organic carbon for providing energy. In this respiratory process, oxygen is consumed along with metabolic substrates while ATP, water, and carbon dioxide are generated. While the energetic payoffs of this process (38 moles of ATP per mole of glucose) are critical to the existence of most multi-cellular life, the risk we oxygen consumers must adapt to is the production of some toxic intermediates as this process occasionally goes astray leading to production of ROS [3]. Under normal conditions, oxygen is reduced to H2O in the myocardium via two paths. Mitochondrial electron transport reduces 95% of O2 by tetravalent reduction to H2O without any free radical intermediates [4]. However, the remaining 5% of oxygen is reduced via the univalent pathway in which free radicals are produced (Fig. 1). When oxygen accepts an electron the superoxide (O2) anion is formed. Superoxide is in equilibrium with its protonated form, HO2. During acidosis, as occurs in ischemia, the protonated form HO2 is favored and more reactive. HO2 is capable of oxidative injury to fatty acids and cell membranes, which is prevented within the cell by dismutation (primarily by superoxide dismutase) to H2O2. H2O2 is only directly toxic at high concentrations not thought to occur under normal conditions in cells. Cells have a well-adapted method to continue the reduction of H2O2 to H2O via catalase or the glutathione system. In this manner superoxide, a byproduct of respiration, is safely metabolized to water. Our tissues have substantial ability to tolerate ROS under normal conditions. However, the setting of ischemia and then reperfusion alters this carefully orchestrated homeostasis. With ischemia antioxidant defenses are eroded and a new danger exists as elevated H2O2 becomes increasingly capable of generating the destructive hydroxyl radical (OH) [5]. Metal ions, particularly iron, may play a role in generating hydroxyl radicals, hence the rationale for metal chelation during oxidative stress [5,6]. Hydroxyl radicals are extremely reactive and may cause direct cell membrane damage, lipid peroxidation, and damage to proteins and sulfhydryl bonds [7]. Additional oxygen-related free radicals (nitric oxide, peroxynitrate, etc.) can also be produced and have important biological effects that can be likewise destructive or protective (but are beyond the scope of this review).

Fig. 1

Reactive oxygen metabolism. Under normal conditions ∼5% of respired oxygen is metabolized to water via this path. Superoxide (O2) and hydrogen peroxide (H2O2) are produced normally and may have protective actions via signaling for preconditioning or oxidant stress induced gene products that activate multiple groups of protective proteins. However, with ischemia and reperfusion the normal balance is lost and hydoyxl radicals (OH) can be produced via the Fenton reaction.

3 Reperfusion injury: reoxygenation gone astray

A large body of experimental literature supports the notion of free radical-generated reperfusion injury when oxygen is reintroduced to ischemic tissue [5,8–13]. This oxidative stress mediated reperfusion injury has been reported in heart, kidney, liver, lung, and intestine. Most of this scientific data is indirect as it is derived from the observation that free radical scavengers improve some aspect of reperfusion injury. For example, Ganote et al. added DMSO, a free radical scavenger, to the reperfusate of ischemic rat hearts and demonstrated decreased CPK release, diminished contraction band formation, and preserved myocardial architecture [14]. When superoxide dismutase plus catalase were administered during reperfusion in an occluded coronary artery canine model, a reduction in infarct size could be seen [15]. However, the protective effect was seen only when the scavengers were administered 15 min prior to reperfusion; no effect could be seen if the drugs were given 15 min after reperfusion. More direct measurements of free radicals are becoming available and although they are technically challenging provide additional evidence for oxidant induced reperfusion injury. Using electron paramagnetic resonance, Zweier et al. has identified a spectral peak in hearts compatible with an oxygen-centered free radical which peaked during the first 10 s of reperfusion [16]. In another experiment, Ambrosio et al. observed that the peak for free radical generation of both carbon- and oxygen-centered radicals occurred rapidly at 15–20 s following reperfusion [17]. He was able to prevent the oxygen-centered peak with superoxide dismutase. Direct hydroxyl radical generation in cardiomyocytes has been shown to peak within 15 min, but this peak of hydroxyl radical did not coincide with cell damage: it may not be the most damaging free radical species [18].

At the isolated cellular level, work from our laboratories using contracting cardiomyocytes demonstrates a pattern of cell death also suggestive of a true reperfusion injury [19–24]. We studied timing of cell death during ischemia and reperfusion in chick cardiomyocytes using propidium iodide staining to determine cell membrane rupture and cell death [20]. Cells were exposed to simulated ischemia with simultaneous hypoxia, hypercarbic acidosis, hyperkalemia, and substrate deprivation, for 1 h followed by 3 h reperfusion (Fig. 2). At the end of 1 h of simulated ischemia 4% of observed cells lost viability; but accelerated cell death was clearly seen immediately after reperfusion. After 3 h of reperfusion the cell PI staining increased to 73% (Fig. 2). By contrast, 4 h of sustained ischemia without reperfusion resulted in PI staining of only 2% after 1 h but only 17% after 4 h. Thus, significantly less cell death as measured by PI occurred during sustained ischemia versus ischemia/reperfusion. Changes in LDH release were consistent with these results. Lengthening ischemia from 30 to 90 min increased ischemia/reperfusion injury as expected, but for any period of ischemia tested over 90% of the accumulated cell death was observed during the reperfusion period not during the ischemic conditions. These results indicate that, although cell injury during ischemia occurs, most of the loss in cell viability occurs during reperfusion. Antioxidants (including metal-chelating agents and thiol donating groups) can substantially protect against this injury, when these agents were administered together in our cardiomyocyte system they reduced cell death after 3 h of reperfusion from 68% to 23% (Fig. 3). We observe that an alteration of reperfusion conditions with antioxidants may significantly improve the survival of cells following ischemia/reperfusion consistent with oxidant mediated reperfusion injury.

Fig. 3

Antioxidant protection during ischemia and reperfusion. Antioxidants 2-mercaptopropionyl glycine and 1,10-phenanthroline given throughout ischemia and reperfusion reduce cell death as measured by propidium iodide uptake after 3 h of reperfusion from 68% (●) in control cells to 23% (■) with the two antioxidants. No difference is seen after only 1 h of ischemia (from Vanden Hoek et al. [21]).

Fig. 2

Cell death pattern suggestive of reperfusion injury in cardiomyocytes. Cardiomyocyte exposed to 1 h of simulated ischemia followed by 3 h reperfusion demonstrate accelerated death during reperfusion not during ischemia. Sustained ischemia (■) for 4 h reveals only 17% cell death compared with 73% cell death in the reperfused (●) cells. A similar pattern is seen for LDH release (from Vanden Hoek et al. [20]).

Further indirect cellular evidence to implicate ROS in reperfusion injury is the “burst” of ROS observed by several prior investigators [25,26]. This burst is demonstrated clearly in cardiomyocytes in the first several minutes of reperfusion using a fluorescent probe sensitive to oxidants [22]. The fluorescent oxidant probe 2′,7′-dichlorofluorescin diacetate is useful (although not highly specific) as an indicator of hydrogen peroxide generation. The probe can enter the cell where it is cleaved by intracellular esterases and trapped as the non-fluorescent dichlorofluorescin (DCFH) molecule. Subsequent oxidation yields the fluorescent product DCF which can be serially measured. Under conditions of ischemia and reperfusion a dramatic “burst” of DCF fluorescence is seen within the first several minutes of reperfusion (Fig. 4). Antioxidants along with several other cardioprotective measures (including ischemic preconditioning and hypothermia) have been observed to diminish this reperfusion burst. This burst of fluorescence appears to peak within 5 min following reperfusion suggesting how rapidly reperfusion events may progress.

Fig. 4

The reperfusion burst of ROS. After 1 h of ischemia a burst of ROS is seen in cardiomyocytes that peaks within 5–6 min as measured by DCF fluorescence. DCF fluorescence is more specific for hydrogen peroxide than superoxide generation, however, also reacts with other oxidants (adapted from Vanden Hoek et al., Ref. [22]).

4 How can cells generate reactive oxygen species under conditions of ischemia and lowered O2?

Despite the very sensible notion that ROS are produced primarily with the reintroduction of oxygen following ischemia, several investigators began to also observe ROS generation during ischemia [27]. While seeming paradoxical at first, there is much literature to support this observation [5]. While a 90% reduction in oxygen delivery would render the heart ischemic (i.e. insufficient oxygen delivery to meet metabolic demands), considerable molecular O2 would still be present. Total anoxia is unlikely to exist even with clinically important ischemia. With ischemia the respiratory cytochromes become redox-reduced allowing them to directly transfer (i.e. “leak”) electrons to oxygen [28]. A redox-reduced cell in the presence of molecular oxygen appears capable of producing large amounts of superoxide anions [28].

The concept that ischemia causes generation of ROS is of major importance because these ischemia-generated ROS appear to play an important signaling role [29], may contribute to direct cellular oxidant damage, and are likely to be the same source of ROS that has been reported to trigger preconditioning [23]. We can observe ischemia-generated oxidant generation in cardiomyocytes and our studies using inhibitors suggest that the mitochondria is the source of these ischemia-generated oxidants [19]. These studies employ the fluorescent probe dihydroethidine (DHE) to observe generation of superoxide in cells exposed to ischemia and reperfusion. DHE enters the cell and is oxidized by ROS, particularly superoxide, to yield fluorescent ethidium which binds to DNA (Eth-DNA), further amplifying its fluorescence. Eth-DNA fluorescence is generally stable, but can be decreased by hydroxyl radical attack. Thus, changes in fluorescence reflect DHE oxidation to Eth-DNA and suggest superoxide generation (see Ref. [22] for ROS specificities of DHE for superoxide rather than H2O2). In cardiomyocytes exposed to simulated 1 h of ischemia, the generation of superoxide is suggested as fluorescence increased from 0.7±0.1 a.u. to a peak of 2.3±0.3 a.u. (Fig. 5A). In addition, various metabolic inhibitors were administered along with ischemia in the presence of DHE. Addition of myxothiazol (0.6 μM), to inhibit electron transport at site III, significantly attenuated peak Eth-DNA fluorescence during 1 h of ischemia from 2.3±0.3 a.u. in controls to 1.3±0.1 a.u. in treated cells (P<0.05, Fig. 5). The addition of site I mitochondrial inhibitors, amytal (2.5 mM) or rotenone (10 μM), also attenuated peak Eth-DNA fluorescence during ischemia (to 1.1±0.1 a.u., P<0.01 and 1.0±0.1, P<0.01, respectively, Fig. 5B). Addition of myxothiazol plus amytal attenuated Eth-DNA fluorescence compared with control cells, but failed to further attenuate Eth-DNA fluorescence compared with either myxothiazol or amytal alone (data not shown). By contrast, the mitochondrial site IV inhibitor cyanide (2.5 mM) failed to attenuate ROS production during ischemia (1.6±0.1 a.u. vs. 1.7±0.1 in controls, Fig. 5C).

Fig. 5

(A) Ischemic oxidant generation decreased with site III mitochondrial inhibitor myxothiazol. The site III mitochondrial inhibitor myxothiazol decreases the elevated generation of superoxide during ischemic as detected by oxidized DHE fluorescence ((●) normal cells; (■) cells with 0.6 uM myxothiazol, from Becker et al. [19]). (b) Ischemic oxidant generation decreased with site I mitochondrial inhibitors rotenone and amytal. The site I mitochondrial inhibitors rotenone and amytal decrease the elevated generation of superoxide during ischemic as detected by oxidized DHE fluorescence ((●) normal cells, (■) cells with 10 uM rotenone, (▴) cells with 2.5 mM amytal, from Becker et al. [19]). (c) Ischemic oxidant generation unchanged with site IV mitochondrial inhibitor cyanide. Addition of site IV inhibitor cyanide (2.5 mM) fails to attenuate DHE oxidation. Collectively, site I and III inhibitors attenuate oxidation, but site IV inhibition fails to attenuate DHE oxidation.

The site of ROS generation along the electron transport chain is suggested by these results with mitchondrial inhibitors and would suggest that oxidant generation must be distal to site I but proximal to site IV. All these inhibitors (amytal, rotenone, myxothiazol, and cyanide) block mitochondrial electron transport, but only those acting upstream of the ubisemiquinone site were able to attenuate the oxidant signal, indicating that the response is not a non-specific response to a blockage of mitochondrial phosphorylation or electron transport. Collectively the findings suggest that mitochondrial ubisemiquinone functions as the primary source of ROS generation during ischemia in this model.

5 Mitochondrial inhibitors attenuate ischemia generated oxidants but not the oxidants seen upon reperfusion: a different intracellular oxidant source?

While we see a pattern of mitochondrial ROS generation (mostly from site III-see above) during the conditions of ischemia, the source(s) responsible for the burst of ROS upon reperfusion are not so clear. Evidence is growing that the mitochondrial source of oxidants seen during ischemia may not be the source of oxidants during the burst of oxidants observed at reperfusion.

In studies that attempt to suppress oxidant generation upon reperfusion (i.e. studies to decrease the “burst” of oxidants seen in Fig. 4) a different picture emerges than results obtained on oxidant generation during ischemia (seen in Fig. 5A–C above). Using DCF fluorescence to detect H2O2 we tested whether mitochondrial inhibitors would also diminish the ROS burst we generated at reperfusion after 1 h of simulated ischemia. Results showed that this burst was not attenuated by the mitochondrial inhibitors amytal, rotenone, myxothiazol, or thenoyltrifluoroacetone as was seen above for ischemia.

Clearly the pattern of inhibitors that decrease oxidant generation during ischemia is different than the pattern seen with the reperfusion peak. The exact source of this reperfusion peak remains uncertain and is the focus of much ongoing research in our laboratory. The change in redox state as the cell transitions from ischemia to reperfusion may also be an important factor in understanding this reperfusion physiology and in designing therapies.

6 ROS-induced ROS release: a positive feedback ROS amplification loop

An important additional concept is the notion of ROS-induced ROS release described by Zorov et al. [30]. Working with isolated adult rat cardiomyctes, they created “triggering” or “inducing” ROS (likely to be singlet oxygen or superoxide anions) via intracellular photoactivation of tetramethylrhodamine compounds. These triggering ROS were associated with mitochondrial depolarization along with mitochondrial permeability transition induction. Observed simultaneously with mitochondrial permeability transition induction was a large burst of ROS from individual mitochondria; suggesting a positive feedback loop of “ROS-induced ROS release” [30]. Supporting this concept was the observation that bongkrekic acid, a mitochondrial permeability transition inhibitor, prevented the large burst of ROS release in a dose–response manner despite the presence of the triggering ROS. This suggests that triggering ROS alone are not sufficient for ROS release rather involvement of the mitochondrial permeability transition may also be required. The mechanism for ROS-induced ROS release remains uncertain but the authors speculate that the triggering ROS induce mitochondrial permeability transition which results in an alteration of mitochondrial membrane fluidity and rigidity which alters the rate of protein–protein interactions required for efficient electron transport. Thus triggering ROS leads to electron transport inhibition which redox reduces the respiratory electron transport complexes which then pass electrons to molecular oxygen resulting in superoxide formation.

7 Free radicals and oxidants also have protective effects

The notion that free radicals and ROS were toxic (and, therefore, primarily injurious) has existed since the early work that noted the similarity of effects of oxygen and ionizing radiation on the body and its tissue [31,32]. However, additional work also suggest the very important role in protection that oxidant signaling may also play within the cell. With these studies comes the current concept that oxidants can also be protective—not simply injurious. Much insight into this came out of work in preconditioning. In the late 1980s Murry et al. described ischemic preconditioning wherein a brief non-lethal episode of ischemia conferred both short and longer-term protection to tissues against an ensuing lethal ischemic insult [33,34]. Preconditioning pathways have been the subject of many studies, and still offers hope that by understanding this natural adaptive protective mechanism new therapies for ischemia will be developed [35]. However, another important lesson is contained (and was almost overlooked) within this seminal description of preconditioning, that antioxidants abolished the preconditioning protection. In other words, treatment with antioxidants interfered with preconditioning protection and made the injury worse.

Vanden Hoek et al. demonstrates the loss of preconditioning protection with antioxidants clearly in cardiomyocytes [23,24]. Isolated cardiomyocytes demonstrate significant preconditioning protection with exposure to a 10-min ischemic preconditioning trigger just before 1 h of ischemia and reperfusion. In these studies, cell death (measured with PI exclusion dye) in non-preconditioned was cells was 47% versus 14% in preconditioned cells (Fig. 6). Oxidant generation was observed to occur in during the brief 10 min of preconditioning ischemia (Fig. 7) that could be attenuated with antioxidants (2-mercaptopriopionyl glycine) and mitochondrial inhibitors (myxothiazol). When 2-mercaptopriopionyl glycine was added only during the preconditioning period in an effort to inhibit this oxidant generation only during the preconditioning stimulus, the protection of preconditioning prior to ischemia was lost (Fig. 8). Under these conditions, increased cell death is the result of adding an antioxidant—that has been protective under other conditions. This observation has been confirmed in several other laboratories that have reported how preconditioning can be produced by ischemia induced oxidants or by direct infusion of the oxidant peroxynitrate and that anti-oxidants (DMSO or MnTBAP) when added during the preconditioning phase interfere with the induction of preconditioning [27,36]. Likewise the “triggering” role for ROS in the induction of late precondition has been established in the whole animal [37]. The addition of 2-mercaptopropionyl glycine was able to abrogate late preconditioning while exogenous oxidants were able to induced late preconditioning in conscious rabbit infarction model, thus confirming the central and protective roles of ROS in protection pathways in animals [37]. The signaling pathways that connect the triggering ROS to induction of preconditioning protection have been the subject of much investigation. A central role in this protection pathway has been established for the mitochondrial ATP-sensitive K channel with opening of the channel cardioprotective during ischemia and reperfusion [38]. Following mitochondrial K-channel opening, both ROS and NO appear to be generated in isolated cardiomyocytes which leads to the cardioprotected or adapted state [39]. Thus, we must appreciate the beneficial effects of ROS. The clinical significance of this physiology and the protective effects of ROS is uncertain but likely to be important in designing human trials.

Fig. 8

Loss of preconditioning protection and increased cell death with antioxidants. When antioxidant 2-mercaptopropionyl glycine (400 uM) is delivered only during the preconditioning period (then washed out prior to ischemia/reperfusion), the protection afforded by preconditioning is lost. Also, hydrogen peroxide (15 uM) alone given in place of hypoxic preconditioning for 10-min results in protection. Thus we demonstrate an ROS mechanism for “paradoxical” cardiomyocyte protection with oxidants and increased cell death due to antioxidants (from Vanden Hoek et al. [23]).

Fig. 7

Oxidant generation during the 10-min preconditioning triggering induction phase. Significant oxidation of DCFH to DCF is seen during the sub-lethal 10-min exposure to hypoxia that results in precondition protection. This oxidation could be completely attenuated by antioxidant 2-mercaptopropionyl glycine (400 uM, from Vanden Hoek et al. [23]).

Fig. 6

Preconditioning protection in cardiomyocytes. A brief (sub-lethal) 10-min exposure to hypoxia produces significant protections from 1 h of ischemia followed by 3 h of reperfusion. Cell death after the 3 h of reperfusion is reduced from 47% to 14% in preconditioned cells (from Vanden Hoek et al. [23]).

8 The central role for oxidant signaling in the cell

The continuous generation of ROS within our tissues is now recognized as a central signaling mechanism for a vast range of metabolic pathways and networks. We have reviewed evidence (above) that oxidant signaling can be used to protect cells by signaling preconditioning protection. Living organisms have not only adapted to protect against ROS, they have developed mechanism for the beneficial uses of free radicals [3]. ROS signaling is important in health as well as under conditions of ischemia. It is not surprising that one must be careful when altering antioxidant defenses. For example, Bai and Cederbaum created a stable transfection of HepG2 cells that over expressed mitochondrial catalase and these cells were indeed more resistant to hydrogen peroxide and antimycin-induced oxidant stress [40]. However, these cells also developed increased sensitivity to tumor necrosis factor-induced apoptosis due to a redox change in the mitochondria. This highlights the careful balance and control mechanisms evolved within our tissues in our oxygen rich environment. ROS are likely to play an important regulatory role in energy production, fertilization, survival kinases activation, ion channel regulation, apoptosis signaling, preconditioning, necrosis, oxygen sensing, inflammatory system, redox homeostasis, and regulation of vascular tone [3]. Some of these important cellular functions are under the control of oxidant inducible genes that code for transcription factors like oxyR or soxR which have been reported to signal entire families of stress response proteins [41]. These induced proteins include antioxidants like Mn-SOD, catalase, glutathione reductase, as well as other protective systems such as NFkB, insulin receptors, glutaredoxin, hydroperoxidase I, alkylhydoperoxide reductase, heat shock proteins, and others; more are currently being described [3,42]. The emerging picture is that cells maintain a delicate balance between the protective oxidant signaling versus detrimental effects and this balance seems a critical aspect of aerobic life.

9 Animal and whole organ models reveal contradictory results

While the evidence is quite strong that significant reperfusion injury can be demonstrated in cellular systems of simulated ischemia and reperfusion, the evidence for reperfusion injury at the whole organ or whole animal level remain contradictory with both positive and negative studies. Jolly et al. reported in 1984 that superoxide dismutase plus catalase administered during early reperfusion could reduce infarct size in dogs from 44% in controls to 22% when infarct size was determined 24 h after ischemia [15]. However, 4 years later Miura et al. reported that no significant benefit was seen from superoxide dismutase plus catalase when infarction in rabbits was assessed 3 days after ischemia [43]. Howitz et al. reported that administration of 2-mercaptopropionyl glycine, a diffusible thiol antioxidant, during reperfusion reduced infarct size by 60% in dogs [43], but Miki et al. found no protection with the same antioxidant even when administered during reperfusion for 4 h in rabbits [44]. The discordant results of ROS scavenging on infarct size have been well reviewed by Downey and Yellon [45]. At the core of this conundrum is to understand whether the acceleration in injury parameters seen during reperfusion is truly the reperfusion-induced death of potentially salvageable cells or whether the injury during ischemia is such that it will lead inevitably to death which merely becomes manifest during the reperfusion period. Tantalizing data from hypothermia would suggest that some portion of the injury is preventable even after ischemia. Miki et al. tested the ability of hypothermia (32 °C) in a 30-min rabbit coronary occlusion model and reported a decrease in infarct size from 51% to 29% even when cooling was only started 10 min prior to reperfusion [46]. Likewise two human trials have reported improved neurological and survival status after cardiac arrest, when cooling was only initiated after return of pulse [47,48]. These whole animal and human studies suggest that at least some portion of organ injury and death is potentially reversible, not inevitable after the ischemic period, thus providing support for the notion of some degree of true reperfusion injury. These contradictory studies raise many important questions on: differences in physiology by species, techniques for determination of infarct size, optimizing pharmacological agents or combinations of agents administered, proper length of time in models of ischemia and reperfusion for observation of protection, most appropriate animal model for infarction, measurement of antioxidant effectiveness, measurement of free radicals, and on differences between buffered perfusate and blood perfusion with white cells plus inflammatory mediators. While the cellular data reveals mechanisms for true reperfusion injury under conditions of simulated ischemia and reperfusion, whether or not the same lethal reperfusion injury is present in whole animals remains uncertain.

10 Clinical uses of antioxidants

Despite many studies and a wide array of antioxidant agents, there are as yet no clinical indications for the routine use of an antioxidant in the setting of ischemia and reperfusion. It is not for lack of trying as multiple clinical trials have attempted to use antioxidants in a wide variety of settings. Flaherty et al. studied the administration of human SOD in the setting of acute myocardial infarction patients who were undergoing percutaneous translumenal coronary angioplasty and found no benefit [49]. A large trial of Vitamin E and beta-carotene likewise failed to show and protective cardiovascular effects when smokers with acute myocardial infarction were treated long term with these agents [50]. In a 2003 recommendation paper on the need for antioxidants in diet an expert panel came to the non-specific conclusion that we should “eat more fruits and vegetables” [51]. There have been positive studies using antioxidants, but the data is not yet compelling. If oxidants play such an important role in reperfusion injury and cardiovascular disease, why is the clinical data so weak at this time?

Several possible explanations may offer insight into the failure of clinical studies of antioxidants. Some experts have suggested that more than one antioxidant is required for clinical effectiveness. The rationale is that antioxidants exist as a “network” wherein both lipid soluble (like tocopherals) and water soluble (ascorbate, glutathione, dihydrolipoic acid) molecules work in a network for the removal of oxidant stress plus the regeneration of oxidant defenses [52]. Work in the isolated rat heart revealed that both vitamin E and dihydrolipoic acid have synergistic functional effects during ischemia and reperfusion [53]. To date clinical trials have generally not used synergistic combinations of agents despite the theoretical advantages and basic science demonstrations of effectiveness.

Another cited explanation for the failure of clinical antioxidant studies is that by inhibiting the normal production of ROS, another “toxic” condition may be produced if these agents or by-products lead to a reduced oxidative phosphorylation and ATP production [54]. For example, the beta-carotene cleavage products have been shown to strongly inhibit state 3 respiration in isolated liver mitochondria [54]. Persistent decreased ATP production has been reported and may be an important consideration following ischemia.

11 Conclusions

Collectively, our cellular and basic data suggest three important aspects for successful cardioprotection: (a) the agent must gain entry to the target tissue and cells, (b) the agents must be delivered in a time sensitive fashion, and (c) the agents must be targeted to the relevant ROS-producing or ROS-consuming intracellular compartment or site (e.g. mitochondria and/or lysosomes etc.).

We know that ischemia results in severely diminished blood flow. How well do our current antioxidants gain entry to the target cells if they are delivered via the blood stream? Because most ischemic emergencies are due to lack of blood supply the ability to get drugs to ischemic tissues must be (almost by definition) quite limited. Current studies have not evaluated this important variable. We see from several investigators the extremely rapid course of reperfusion, with oxidant generation peaking within 5 min or less. Can we reasonably expect our current trial design to treat patients in this time window? Importantly most basic science studies that show protection use pretreatment (which is impossible in most clinical settings) or treat very rapidly with the first minute of reperfusion. The lessons of basic science would suggest that even fairly brief delays in reperfusion therapies are associated with loss of protection. Moreover, we are now identifying specific ROS species, sites of ROS generation, and intercellular compartments that appear important for selective targeting by our therapies, it will not be surprising if therapies fail if they do not successfully enter these compartments or treat these specific ROS species.

What about the potential for administered antioxidants to decrease the protective response in some tissue? It is possible that our current antioxidant compounds have failed in clinical trials because they made some tissue better but also contributed to additional injury in other adjoining tissues. Infarction is not a homogenous insult, some areas are profoundly ischemic while others are less so. Over time the border zone of ischemic injury may expand. It is possible that cells in the most ischemic portion of the heart are protected by antioxidants while adjoining tissues are undergoing some element of ischemic preconditioning signaling and adaptation. These nearby tissues may be injured by the non-specific addition of antioxidants which interfering in adaptive natural cardioprotection. As our mechanistic understanding grows in the coming years it is very likely that clinical therapies will evolve and become more important. However, therapies must become more specific for site of action, not have deleterious effects on other signaling pathways, be targeted to a specific ROS or cellular compartment, and the time sensitive challenge of delivering the correct therapy at precisely the correct time must be surmounted for us to predict reliable clinical success.


  • Time for primary review 29 days


  1. [1]
  2. [2]
  3. [3]
  4. [4]
  5. [5]
  6. [6]
  7. [7]
  8. [8]
  9. [9]
  10. [10]
  11. [11]
  12. [12]
  13. [13]
  14. [14]
  15. [15]
  16. [16]
  17. [17]
  18. [18]
  19. [19]
  20. [20]
  21. [21]
  22. [22]
  23. [23]
  24. [24]
  25. [25]
  26. [26]
  27. [27]
  28. [28]
  29. [29]
  30. [30]
  31. [31]
  32. [32]
  33. [33]
  34. [34]
  35. [35]
  36. [36]
  37. [37]
  38. [38]
  39. [39]
  40. [40]
  41. [41]
  42. [42]
  43. [43]
  44. [44]
  45. [45]
  46. [46]
  47. [47]
  48. [48]
  49. [49]
  50. [50]
  51. [51]
  52. [52]
  53. [53]
  54. [54]
View Abstract