INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

REACTIVE OXYGEN SPECIES (ROS) have been proposed to play both physiological roles in cell signaling in skeletal muscle (13) and pathological roles in the skeletal muscle damage and degeneration that occur in a number of different situations (12). A considerable amount of data has been presented concerning the role of these species in the muscle damage that accompanies ischemia and reperfusion or unaccustomed or excessive exercise (5, 7, 12, 14, 17, 19). Skeletal muscle is recognized to be relatively resistant to injury due to ischemia, but there are important clinical examples of where this damage does occur, such as after prolonged use of a tourniquet in orthopedic surgery or after surgery to correct arterial occlusion (10, 16). Several investigators [e.g., Oredsson et al. (15) and Perler et al. (18)] have reported that administration of scavengers of free radical species reduced reperfusion injury to skeletal muscle, and we have confirmed that this protection is apparent with ascorbate but not with some other potential scavengers (5). Our previous data indicate that reperfusion of ischemic skeletal muscle was associated with an ~75% increase in hydroxyl radical activity in the interstitial space, which was speculated to contribute to the oxidation of muscle glutathione, protein thiols, and lipids also seen in this model. No evidence for release of superoxide into the interstitial space was found during reperfusion (17).

Previous studies of exercise have demonstrated the presence of increased activities of superoxide (13) and hydroxyl radicals (14) in muscle extracellular fluid and the appearance of a more stable free radical species in the circulation of exercising humans (1). Ashton et al. (1) used used -phenyl-tert-butylnitrone (PBN) as a spin trap and electron-spin resonance (ESR) to demonstrate an increase in the content of a free radical species in the blood of exercising subjects. The free radical species was not fully identified, but comparison of the hyperfine coupling constants with published data (8, 20) indicated that the circulating species detected were likely to be carbon- or oxygen-centered lipid radicals (2).

Because ischemia and reperfusion of skeletal muscle are associated with an increase in release of ROS (15) and biologically active lipids, such as prostanoids (11), to the extracellular fluid, we examined whether ischemia and reperfusion also lead to generation of the ESR-detectable lipid free radical species in the circulation. In addition, to investigate further the site and mechanism of formation of the free radical species that is detectable by ESR, we examined whether this species is found in the interstitial fluid after ischemic-reperfusion injury to skeletal muscle and the relationship of the ESR-detectable signal to interstitial superoxide levels, hydroxyl radical activity, and prostanoid content.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Ischemia-reperfusion of the rat limb. The model of ischemia-reperfusion used was developed by Klenerman et al. (10) and modified by McArdle et al. (11) to examine ischemia-reperfusion injury in the rat. Experiments were carried out in accordance with UK Government Home Office guidelines under the UK Animals (Scientific Procedure) Act, 1986. Adult female Wistar rats (~200 g) were anesthetized with pentobarbitone sodium (65 mg/100 g ip) throughout the procedure. Microdialysis probes (MAB 3.8.10, 10-mm membrane, 0.5-mm wide, Metalant; Stockholm, Sweden) with a molecular mass cutoff 35,000 Da were placed into the tibialis anterior of both limbs using a splittable 22-gauge plastic introducer. Probes were perfused with either 140 mM PBN (Alexis Biochemicals; Nottingham, UK) in normal saline, 0.5 mM salicylate in normal saline at a flow rate of 1 µl/min, 50 µM cytochrome c in normal saline at a flow rate of 4 µl/min, or normal saline alone and allowed to stabilize for 30 min. The femoral artery of one leg was exposed and occluded using a microvessel clip for 4 h. During this period, samples were collected from the probes over sequential 60-min periods. After collection of the sample immediately preceding reperfusion, the microvessel clip was removed, and the ischemic limb was allowed to reperfuse for a further 60 min, with one sample collected. At the end of 60 min of reperfusion, 1 ml of arterial blood was obtained from the aorta, and animals were euthanized with an overdose of anesthetic. Blood samples were also collected from a group of control nonischemic animals on euthanization.

Analysis of blood and microdialysates by ESR. The blood samples obtained on euthanization of the animals were immediately mixed with an equal volume of PBN (140 mM) in saline. The mixture was allowed to stand on ice and then mixed with an equal volume of toluene, and the PBN adduct was extracted and stored at 80°C until analyzed by ESR. Microdialysates containing the nitroxide spin trap PBN were added to an equal volume of toluene, mixed, and allowed to separate into two layers. The top layer (containing the PBN adduct) was removed and stored at 80°C until analyzed.

3

Hydroxyl and superoxide radical analyses. 2,3-Dihydroxybenzoic acid (2,3-DHB) generated from the hydroxyl radical reaction with salicylate in the microdialysis fluids was measured as an index of hydroxyl radical activity by HPLC with electrochemical detection as previously described (17). The reduction of cytochrome c in the microdialysate was used to calculate superoxide in microdialysates (13).

Prostaglandin analysis. The PGE₂ content of 20-µl samples from the saline microdialysates was analyzed using a competitive-binding radioimmunoassay with a rabbit anti-PGE₂ antibody (DuPont NEN Biochemicals) as previously reported (11).

Statistical analyses. Data are presented as means ± SE of values from four to six animals for each experimental condition. Data were initially analyzed by repeated-measures ANOVA. Where significance was indicated, means were compared using the Bonferroni correction.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

ESR analysis of the blood sample obtained 1 h after reperfusion of the ischemic limb revealed the presence of a PBN adduct (Fig. 1B). In comparison, the species was either undetectable or present at very low levels in the blood of control anethetized animals (Fig. 1A). The signal in the blood of rats subjected to limb ischemia and reperfusion had a g value (a proportionality constant between the magnetic and angular momentum of the free radical) and hyperfine splitting constants consistent with those previously reported from the PBN-trapped species detected in human blood (1).

View larger version (23K): [in this window] [in a new window]   Fig. 1.   Examples of electron-spin resonance (ESR) spectra of -phenyl-tert-butylnitrone (PBN)-trapped adduct in blood of control anesthetized rats (A) and anesthetized rats at 60 min after reperfusion of the ischemic tibialis anterior muscle (B).

The same PBN adduct was also observed in the microdialysates from the PBN-perfused probes, indicating the presence of the same species in muscle interstitial fluid (Fig. 2). A significant increase in the amplitude of the signal was seen at both 0-2 and 2-4 h of ischemia, and this was further increased by a mean of 420% during reperfusion (Fig. 3 and Table 1). Signal intensity in perfusates from the contralateral control limb also appeared to increase during the procedure, but no significant changes were observed. Figure 2, A and B, shows typical example spectra for PBN-trapped adducts in microdialysates. Additional features were also apparent in the spectra from some of the preischemic (Fig. 2A) and reperfused (Fig. 2B) microdialysis samples; however, these were not present in microdialysis samples from all animals. The origin of these signals is unclear, and they may be artifactual, a possibility that is currently under study.

View larger version (22K): [in this window] [in a new window]   Fig. 2.   Typical ESR spectra of PBN-trapped adducts in microdialysates from the test tibialis anterior muscle before initiation of ischemia (A) and during reperfusion after 4 h of ischemia (B).

View larger version (14K): [in this window] [in a new window]   Fig. 3.   Amplitude of the ESR signal from PBN-trapped radical species in microdialysates from ischemic-reperfused (open bars) and contralateral (gray bars) limbs of anesthetized rats. *P < 0.05 compared with nonschemic control muscles and preischemia values.

2,3-DHB, the product of oxidation of salicylate by hydroxyl radicals, was detected in all microdialysate samples from the tibialis anterior muscle (Table 1). In the microdialysates of the ischemic limb, levels of 2,3-DHB increased significantly during ischemia and further increased during reperfusion. The levels from the contralateral limb also showed some significant increases during both the ischemic and reperfusion periods in the test limb, but these were less than in the test limb. Superoxide levels in the microdialysates from the ischemia-reperfused limb did not differ from the contralateral limb throughout the study (Table 1). PGE₂ levels were found to increase significantly during the reperfusion period (Table 1).

Linear correlations were undertaken between the PBN adduct in microdialysates with the 2,3-DHB, superoxide, and PGE₂ contents and are shown in Table 2. These demonstrate a significant positive linear correlation between the PBN adduct and the 2,3-DHB content (r = 0.815) of microdialysates and a less strong but significant correlation with the PGE₂ content (r = 0.730). No correlation was seen between the PBN adduct and the microdialysate superoxide content (r = 0.14). The r² values for the relationships between the PBN adduct and the 2,3-DHB and PGE₂ were 0.664 and 0.532, respectively, indicating that >50% of the variation in the magnitude of the PBN adduct could be explained by the changes in each of these other two variables.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The data presented here indicate that reperfusion of ischemic skeletal muscle is associated with the presence of a free radical species in blood that is trapped by PBN. The species appears to be very similar or identical to that seen by Ashton et al. (1) in human serum after exhaustive exercise. The hyperfine coupling constants (a_N and a_H) from the current spectra are similar to those reported by Ashton et al. (1, 2) (a_N = 13.60 G and a_H = 1.56 G). The nature of the trapped species is not clear, but Ashton et al. (2) suggested that the species observed were either secondary carbon-centered or alkoxyl radicals. Similar species also appear to have been detected by ESR-spin trapping after reperfusion injury to cardiac tissue (8, 9, 20). It is therefore clear that the radical species observed in the circulation is relatively nonspecific, being produced from multiple species (rats or humans), tissue sources (cardiac or skeletal muscle), and origins of the oxidative stress (exercise or ischemia-reperfusion injury). However, while the precise nature of the trapped radical is presently unclear, it is sufficiently stable enough to be detectable by ESR-spin trapping and is likely to be <35,000 Da in molecular mass because this is the cutoff weight for the microdialysis probe. Furthermore, the decreased intensity of the high-field lines on the ESR spectra is suggestive of spin inhibition. One explanation of this is that the adduct is "carrying" a longer-length biopolymer radical, perhaps incorporating a carbon chain of a polyunsaturated lipid.

The ESR-detectable species was also found to be present in microdialysis fluids from skeletal muscle, particularly after ischemia and reperfusion (Fig. 2). These data argue against plasma lipids as a potential source for the species. Zini et al. (21) also showed increased free radical signals detectable by ESR in the microdialysate of rats after 30 min of global ischemia to the rat forebrain. These authors used the nitroxide pyridyl-N-oxide-tert-butylnitrone, which is similar to PBN, as the spin-trapping agent, and examination of the hyperfine coupling constants of the ESR spectra obtained indicates that the species was likely to be a carbon-centered radical.

Garlick et al. (8) and Ashton et al. (1, 2) suggested that the likely route of formation of the circulating ESR-detectable species involved reaction of primary oxygen free radicals (such as hydroxyl and superoxide radicals) with membrane lipids. This mechanism is further supported by the work of Bolli et al. (4), who reported that the hyperfine splitting constants of the signal were consistent with the trapping by PBN of alkoxyl and/or alkyl radicals formed by the peroxidation of polyunsaturated lipids. A comparison of the appearance of the PBN-trapped species with production of 2,3-DHB from salicylate in the microdialysis probe indicated a correlation between these species (Table 2). Because the formation of 2,3-DHB from salicylate is a specific marker of hydroxyl radical activity, these data argue that the rate of production of the lipid radical is regulated by the hydroxyl radical activity. Measurement of the circulating ESR-detectable species may therefore provide a relatively noninvasive technique of assessing extracellular hydroxyl radical activity in skeletal muscle (and potentially other tissues) subjected to oxidative stress, such as ischemia-reperfusion or contractile activity.The correlation between PGE₂ in the microdialysate and the ESR signal additionally suggests that released lipids, such as prostanoids, may be a source of the lipid involved in the generation of the radical species. Prostanoid metabolism involves generation of unstable lipid free radical intermediates, but these would be unlikely to be sufficiently stable to persist in the circulation. In contrast, there was no correlation with superoxide content because the current data confirm our previous observation of no rise in extracellular superoxide after ischemic-reperfusion injury to skeletal muscle (17). Previous studies have demonstrated that an increase in superoxide release could be detected in microdialysates from muscle interstitial fluid during contractile activity (13, 17) and hence imply that the lack of a rise in superoxide after ischemia and reperfusion is not artifactual. Tissues contain extracellular superoxide dismutase that acts to maintain extracelluar superoxide at low concentrations, but whether this enzyme is present in muscle interstial fluid is not known.

A statistically significant increase in hydroxyl radical activity was seen in muscle interstitial fluid from the contralateral nonischemic limb (Table 1), as previously reported (17). There is some evidence that this model of ischemia and reperfusion may be associated with compensatory changes in blood flow in the contralateral control limb (11) that may influence the recovery of 2,3-DHB across the microdialysis membrane and thus the apparent hydroxyl radical activity (17).

Our data therefore indicate the lipid free radical species detectable by PBN trapping with ESR analysis in blood after ischemia-reperfusion is likely to be generated by hydroxyl radical attack on a lipid that is present either within, or adjacent to, the interstitial fluid. This lipid is likely to be polyunsaturated and may be related to lipids, such as prostanoids, released during reperfusion (11), but alternative potential lipid sources including muscle membrane lipids could also be involved. It is currently unclear whether this radical species plays any physiological or pathophysiological role. Although the ESR spectra are compatible with the presence of either carbon- or oxygen-centered lipid radicals, the relatively long-lived nature of the radical is more compatible with a carbon-centered than an oxygen-centered species and hence suggests that alkyl lipid radicals are the more likely source. Preliminary data indicate that in vitro oxidation of -linolenic acid yields a species trapped by PBN with almost identical ESR hyperfine coupling constants to those seen here and further attempts to clarify the nature of the contributing radical species by a variety of analytic techniques are already in progress (G. W. Davidson and T. Ashton, unpublished observations). Ashton et al. (2) also showed that ascorbic acid was an effective scavenger of not only the circulating ESR-detectable species generated after exhaustive exercise, but also lipid peroxides generated by exercise, data that are entirely compatible with the potential mechanism of formation described above.

Both alkoxyl and alkyl lipid radicals would be expected to undergo reactions in vivo that could affect sites distant to the site of origin (in this case, muscle interstitial fluid). These effects would involve oxidation of biological molecules. Unilateral limb ischemia is associated with release of a number of mediators of systemic effects (3, 6). These effects include an apparent increase in oxidative stress in tissues not subjected to the ischemia (17), and it seems likely that the ESR-detectable lipid radical reported here contributes to such systemic effects.

In conclusion, we demonstrated that unilateral limb ischemia of the rat hindlimb leads to formation of a relatively long-lived, ESR-detectable free radical in muscle interstitial fluid. The mechanism of formation of this radical appears to involve hydroxyl radical reaction with a released lipid. The ESR-detectable species persists within the blood of reperfused animals. It may potentially mediate some systemic consequences of unilateral limb ischemia and additionally provide a relatively noninvasive means of assessing hydroxyl radical production within the interstitial space.