Preventing oxygen free-radical injury in ischemic revascularized bone grafts

Superoxide radicals have been shown to play a role in the cellular injury of reperfused ischemic tissues. We examined the protective effect of superoxide dismutase (SOD), a superoxide radical scavenger, on the reperfusion injury of replanted vascularized bone grafts after 4- and 8-hour periods of ischemia in a rat model. Histologic, fluorochrome, and histomorphometric analyses showed no difference between 4-hour superoxide dismutase-treated and control grafts, with both groups appearing viable. Similar analyses of the 8-hour ischemic grafts revealed both a qualitative and statistically significant quantitative difference (p less than 0.001) between the superoxide dismutase-treated and control grafts in parameters related to viability. Our results indicated that the administration of superoxide dismutase to free vascularized grafts by means of intraarterial perfusion after prolonged periods of warm ischemia significantly enhances the survival of these grafts.     

Protective effects of superoxide dismutase against ischemia-reperfusion injury: development and application of a transgenic animal model

Reperfusion of ischemic tissues can be associated with structural and functional injury, which is referred to as ischemia-reperfusion injury. Superoxide dismutase is an endogenous free radical scavenger that converts toxic oxygen derived free radicals to hydrogen peroxide. With the development of gene cloning technology, the potential of manipulating cells to overexpress endogenous proteins has been realized. Transgenic mice capable of overexpressing superoxide dismutase, and knockout mice in which the gene responsible for its production has been deleted, were used as a model to examine the protective effects of superoxide dismutase against ischemia-reperfusion injury. Epigastric island flaps were elevated in wild-type (control), transgenic superoxide dismutase 1, and knockout superoxide dismutase 1 mice and subjected to ischemic intervals of 0, 3, 6, 9, or 12 hours. Five animals were studied at each time point in each study group. Flap viability was assessed on postoperative day 7. Baseline wild-type flap survival was 100 percent after 3 hours of ischemia and subsequent reperfusion; survival decreased to 21 percent after 9 hours of ischemia. Transgenic mice had significantly higher flap survival than wild-type animals after 6 hours of ischemia and subsequent reperfusion (97.0 versus 85.2 percent) and after 9 hours of ischemia (82 versus 21 percent, p < 0.01). In knockout mice, there was complete flap necrosis after as little as 3 hours of ischemia. This study confirms the protective effects of superoxide dismutase against ischemia-reperfusion injury. In addition, its deficiency results in a dramatic susceptibility to ischemic injury.     

Differing flow patterns between ischemically challenged flap skin and flap skeletal muscle: implications for salvage regimens.

In this study, the authors tested the hypothesis that there is a significant difference in spatial patterns of reflow in skin as opposed to skeletal muscle after an ischemic insult. The authors believe that this pathophysiologic difference between the two flap types has significant implications for flap salvage strategies. Bilateral buttock skin flaps (10 x 18 cm) and latissimus dorsi myocutaneous flaps (10 x 20 cm) were elevated in Landrace pigs (n = 7). Flaps on one side of the animal were randomly assigned to 6 hours of arterial occlusion, with the contralateral side acting as control. At 15 minutes, 1 hour, and 4 hours after reflow, radioactive microspheres (15 microm) were injected into the left ventricle. After 18 hours of reperfusion, skin and muscle viability were estimated by intravenous fluorescein and soaking in nitroblue tetrazolium, respectively. Flow rates in the skin with an ischemia-reperfusion injury were significantly reduced (30 to 53 percent), at all time intervals, compared with controls. The flow rate in the fluorescent skin with ischemia-reperfusion injury of the latissimus dorsi flaps (0.037 ml/min/g at 15 min) was greater than in that of the buttock flaps (0.018 ml/min/g). The muscle flaps with ischemia-reperfusion injury had significantly higher flow rates than control muscle flaps at all time intervals studied (at 1 hour, 0.32 ml/min/g compared with 0.16 ml/min/g, respectively). In flap skeletal muscle, an early hyperemic phase during reperfusion maintains a significant blood flow to all regions, including the area of the flap that is destined for necrosis. In flap skin, however, there is a marked decrease in flow rates. These differences have important implications for the intravascular delivery of therapeutic agents to the damaged portions of the flap. Transdermal drug delivery systems should be explored as an alternative to intravascular regimens for the salvage of flap skin with ischemia-reperfusion injury.     

Angiogenic and anti-inflammatory properties of azadirachtin A improve random skin flap survival in rats

Random skin flaps are widely used to repair tissue defects. However, the distal flap regions are prone to ischemic necrosis, limiting clinical applications. Azadirachtin A, a fruit extract from the neem, improves tissue blood supply and metabolism, reduces cell swelling, promotes tissue healing, and prevents venous thrombosis. We explored whether it enhances random skin flap survival. Fifty-four Sprague-Dawley rats were divided into control, low-dose, and high-dose Azadirachtin A-treated groups using a random number table. We used an improved version of the McFarlane technique to create flaps. On day 2, superoxide dismutase and malondialdehyde levels were measured. Tissue slices prepared on day 7 were stained with hematoxylin and eosin. The expression levels of vascular endothelial growth factor (VEGF), toll-like receptor 4 (TLR4), nuclear factor kappa-B (NF-kB), interleukin-1β (IL-1β), interleukin-6 (IL-6), and tumor necrosis factor-α (TNF-α) were immunohistochemically assayed. Microcirculatory blood flow was measured via laser Doppler blood flowmetry. Flap angiography was performed using the lead-oxide gelatin injection technique. And the azadirachtin A groups exhibited a greater mean flap survival area, an improved mean blood vessel density, a greater blood flow, and higher superoxide dismutase and VEGF levels, especially at the high dose. Azadirachtin A markedly reduced the levels of TNF-α, IL-6, IL-1β, TLR4, and NF-kB. These findings suggest that azadirachtin A promotes random skin flap survival by improving the blood supply, reducing tissue inflammation, and inhibiting flap ischemia reperfusion injury.     

Improved survival of island flaps after prolonged ischemia by perfusion with superoxide dismutase

Perfusion of rat groin flaps after 10 and 11 hours of complete ischemia with superoxide dismutase, an oxygen free-radical scavenger, significantly improved the survival of these flaps. This finding provides further evidence for the important role that oxygen-derived free radicals play in ischemic injury. The study also demonstrates that while restoration of blood supply alone is not enough to ensure tissue survival after prolonged ischemia, chemical agents can be utilized to achieve viable flaps beyond what was believed to be "a point of no return".     

Reduction of paraquat toxicity by superoxide dismutase

The effect of intravenously administered superoxide dismutase on paraquat-treated rats kept either in air or an atmosphere of 90%–95% oxygen was investigated. Of those rats maintained in the oxygen-enriched atmosphere, 50% died within 30 hours whereas, 50 hours elapsed before 50% mortality was observed for the superoxide dismutase-treated rats. Those animals allowed to remain in air were more responsive to superoxide dismutase treatment. Of those animals for which paraquat was fatal, untreated rats showed 50% cumulative mortality within 35 hours after paraquat administration, whereas those rats treated with superoxide dismutase showed 50% mortality after 80 hours. Sections of lung tissue examined at low magnification indicated that the extensive alveolar and vascular damage caused by paraquat was ameliorated with the administration of superoxide dismutase. These findings may have particular relevance in the treatment of paraquat intoxication in humans.     

Vascular endothelial growth factor expression in pig latissimus dorsi myocutaneous flaps after ischemia reperfusion injury

Exogenous administration of vascular endothelial growth factor (VEGF) improves long-term viability of myocutaneous flaps. However, endogenous expression of this substance in flaps following ischemia-reperfusion injury has not been reported previously. Endogenous production of VEGF was measured in myocutaneous pig latissimus dorsi flaps after ischemia-reperfusion injury. Latissimus dorsi myocutaneous flaps (15 x 10 cm) were simultaneously elevated bilaterally in six Yorkshire-type male pigs (25 kg). Before elevation, three flap zones (5 x 10 cm) were marked according to their distance from the vascular pedicle. After isolation of the vascular pedicle, ischemia-reperfusion injury was induced in one flap by occlusion of the thoracodorsal artery and vein for 4 hours, followed by 2 hours of reperfusion. The contralateral flap served as a control. Perfusion in each zone was monitored by laser Doppler flowmetry at baseline, during ischemia, and during reperfusion. At the end of the protocol, skin and muscle biopsies of each flap zone and adjacent tissues were obtained for later determination of VEGF protein levels. VEGF concentrations were quantified using the Quantikine human VEGF immunoassay. Skin perfusion was similar among all flap zones before surgery. Flow fell in all flaps immediately after flap elevation. After 4 hours of ischemia, blood flow in the ischemic flaps was significantly decreased (p < 0.05) compared with nonischemic control flaps. After 2 hours of reperfusion, flow in ischemic flap skin recovered to levels similar to those in control flaps. VEGF protein concentrations in muscle tissue exceeded concentrations in skin and decreased from zones 2 to 3 in control and ischemic flaps. No significant differences in VEGF concentrations between ischemic and control muscle zones were observed. However, the concentration of VEGF in all muscle zones was significantly higher (p < 0.05) than muscle adjacent to the flap. Concentrations in skin zones 1 and 2 were significantly higher (p < 0.05) in ischemic flaps than in control flaps, but levels in zone 3 (most ischemic flaps) showed no significant difference.     

Evidence of direct toxic effects of free radicals on the myocardium

The hypothesis that oxygen-derived free radicals do indeed play a role in myocardial ischemic and reperfusion injury has received a lot of support. Experimental results have shown that free radical scavengers can protect against certain aspects of myocardial ischemic injury and that on reperfusion the heart approaches a level that is more normal than those hearts not receiving additional scavenging agents. Superoxide dismutase, catalase, glutathione peroxidase, hydroxyl radical scavengers and iron chelators such as desferrioxamine have proven successful in providing an increased level of recovery. These results indicate, as would be expected, that superoxide, hydrogen peroxide and hydroxyl radicals may all, at some point, either contribute to the injury or be important in generating a subsequent radical which causes damage. In addition, solutions capable of generating free radicals have been shown to cause damage to myocardial cells and the vascular endothelium that is similar to the damage observed during myocardial ischemic and reperfusion injury. Alterations in function, structure, flow, and membrane biochemistry have been documented and compared to ischemic injury. The continued investigation of the role of free radicals in ischemic injury is warranted in the hope of further elucidating the mechanisms involved in free radical injury, the sources of their generation, and in defining a treatment that will provide significant protection against this particular aspect of ischemic damage.     

Increased survival of skin flaps by scavengers of superoxide radical

Elevation of rat abdominal skin flaps, followed by ligation and division of the left inferior neurovascular pedicle, resulted in only a 40% survival of the area normally perfused by the ligated artery and vein. Superoxide dismutase (SOD) (EC 1.15.1.1) administered i.v. (20,000 U/kg) 30 min before flap elevation increased survival to 52%. SOD derivatized with polyethylene glycol, which increases circulating half-life, was more effective, increasing survival to 80%. This protective effect resulted from the catalytic activity of the derivatized enzyme because inactivation by treatment with H2O2 eliminated its effect on skin flap survival. An equimolar mixture of Desferal and MnCl2, which catalyzes the dismutation of O2- in vitro, improved survival to 72%. Desferal-Fe3+, which lacks in vitro SOD activity, or Mn2+ alone did not affect the survival of skin flaps, but Desferal alone was nearly as effective as the Desferal-Mn2+ mixture. This effect of Desferal may result from acquisition of and subsequent removal of iron in vivo. These results support the view that the superoxide radical or a product derived from it plays a role in limiting the survival of island skin flaps.     

Morphologic characterization of acute injury to vascular endothelium of skin after frostbite

Cyclo-oxygenase inhibitors and free-radical scavengers protect the skin against necrosis induced by frostbite. However, the tissue component(s) that determine the evolution of skin necrosis and the mechanism of this pharmacologic protection are not precisely defined. We have studied freezing injury to rabbit ears by serial biopsies examined by light and electron microscopy. The morphologic evidence of skin injury due to freezing was localized exclusively in the endothelial cells, particularly in the arterioles. Within 1 hour, the entire microvasculature demonstrated endothelial damage. Intravascular platelet aggregation occurred just after thawing and closely paralleled the endothelial cell injury. Very few neutrophils were seen initially (at 10 minutes). By 1 hour, leukocyte aggregates were present, and they further increased at 6 hours. Swelling of the interstitium started 10 minutes after thawing, while extravasation of erythrocytes began to appear by 6 hours. Parenchymal elements of skin were relatively free of damage. In the ear cartilage, the chondrocytes showed evidence of damage immediately after freezing. The administration of superoxide dismutase (SOD) during thawing (reperfusion) did not qualitatively alter any of the initial morphologic changes induced by freezing. We conclude that the endothelial cell is the initial target of injury induced by freezing, an initial injury that is mediated by a non-free-radical-mediated mechanism. It is likely that this acute injury ultimately compromises blood flow and leads to skin necrosis.     

Presence and activity of nitric oxide synthase isoforms in ischemia-reperfusion-injured flaps

Nitric oxide is produced from the amino acid L-arginine by nitric oxide synthase, which has three known isoforms: (1) endothelial nitric oxide synthase and (2) brain nitric oxide synthase, both of which are constitutive nitric oxide synthase; and (3) inducible nitric oxide synthase. The authors' hypothesis is that after reperfusion injury, endothelial cell dysfunction leads to disruption of nitric oxide synthase-mediated nitric oxide production and that this may in part explain the deleterious effects of ischemia-reperfusion injury on tissue survival and blood reflow in flaps. An experiment was designed to study the effects of ischemia-reperfusion injury on the bioactivity of all three isoforms of nitric oxide synthase. Buttock skin flaps and latissimus dorsi myocutaneous flaps were elevated in eight pigs. Flaps on one side of the animal were randomized to receive 6 hours of arterial ischemia, whereas flaps on the other side served as controls. At 6 hours of ischemia and at 1, 4, and 18 hours after reflow, tissue biopsy specimens were obtained and were processed for both constitutive nitric oxide synthase and inducible nitric oxide synthase enzyme activity on the basis of the L-citrulline assay. In addition, specimens were processed for Western blot analysis of the three isoforms. The authors' results revealed three key findings: first, there was a statistically significant (p < 0.001) decrease in constitutive nitric oxide synthase activity of ischemia-reperfusion-injured flaps as compared with controls in both skin and muscle for all time intervals measured. Second, Western blot analyses of endothelial nitric oxide synthase and brain nitric oxide synthase showed a significant decrease in the signal intensity in ischemic and reperfused tissue as compared with controls. Third, the inducible nitric oxide synthase isoform's activity and protein remained undetectable in both tissue types for all time points measured. The authors' data demonstrated that following ischemia-reperfusion injury in the pig flap model there was a disruption of constitutive nitric oxide synthase expression and activity, which may lead to decreased nitric oxide production. The significant decrease in nitric oxide synthase activity found in the current study may partly explain the mechanism of tissue damage in flaps subjected to ischemia-reperfusion injury. Knowledge of the kinetics of nitric oxide synthase activity under conditions of ischemia-reperfusion injury has important implications for the choice and timing of delivery of therapeutic agents whose goal is to increase the bioavailability of nitric oxide in reperfused tissue.     

Circulatory and metabolic events in pig island skin flaps after arterial or venous occlusion

After 1 hour of arterial or venous occlusion, the circulatory and metabolic events in island skin flaps of the pig were studied. Both occlusion types showed significant but transient increases in glucose uptake and a parallel release of lactate, hypoxanthine, and potassium. Oxygen uptake and noradrenaline release were not significantly affected. No significant difference between the arterial and venous occlusions was seen in the metabolic parameters. The flap blood flow, measured by total venous outflow and laser Doppler flowmetry, was significantly lower after venous than after arterial occlusion. This long-lasting difference in flow response may help to explain the observation that venous occlusion is more deleterious to skin flaps than arterial occlusion. A mechanism underlying these results may be more pronounced microthrombotization and/or edema formation after venous occlusion than after arterial occlusion.     

Chimeric SOD2/3 inhibits at the endothelial-neutrophil interface to limit vascular dysfunction in ischemia-reperfusion

After an ischemic episode, reperfusion causes profound oxidative stress in the vasculature of the afflicted tissue/organ. The dysregulated accumulation of reactive oxygen species (ROS), such as superoxide, has been closely linked to the production and release of proinflammatory mediators, a profound increase in adhesion molecule expression by the vascular endothelium, and infiltration of neutrophils during ischemia-reperfusion (I/R). Superoxide dismutase (SOD) has been shown to protect tissues and organs against I/R-induced injury; however, the drug had to be continuously perfused or kidneys had to be occluded to prevent clearance. We used intravital microscopy, a system that allowed us to visualize neutrophil-endothelial interactions within the mesenteric postcapillary venules of cats subjected to I/R and tested the hypothesis that I/R-induced neutrophil recruitment was inhibited by treatment with SOD2/3. SOD2/3 is a chimeric fusion gene product that contains the mature SOD2 as well as the COOH-terminal "tail" of SOD3 and, unlike the three naturally occurring SODs (SOD1, SOD2, and SOD3), which bear a net negative charge at pH 7.4, SOD2/3 is positively charged and physiologically stable. Our results suggest that not only does SOD2/3 have a much greater efficacy in vivo than the native human SOD2, but its administration prevents I/R-induced neutrophil-endothelial cell interactions and microvascular dysfunction. Moreover, our data support the hypothesis that reactive oxidants mediate I/R-induced injury and that the chimeric recombinant SOD2/3 has the potential to be a therapeutic agent against this debilitating illness.     

In vivo inhibition of superoxide dismutase in mice by diethyldithiocarbamate.

Superoxide dismutase was assayed by a method which takes advantage of the inhibitory action of superoxide dismutase (or tissues which contain superoxide dismutase) on the rate of autooxidation of 6-hydroxydopamine. Incubation of pure superoxide dismutase of homogenates of brain or liver with 10(-3) M diethyldithiocarbamate for 1.5 hours resulted in total loss of superoxide dismutase activity. Inhibition of superoxide dismutase was not reversed by dialysis, but after dialysis, enzymatic activity was restored with CuSO4. When 1.5 g of diethyldithiocarbamate/kg were injected into mice, the superoxide dismutase activity at 3 hours was decreased by 86%, 71%, and 48%, respectively, in whole blood, liver, and brain. A dose of 0.5 g of diethyldithiocarbamate/kg lowered the superoxide dismutase activity by 42% in liver at 3 hours. A study of the time course for inhibiton of superoxide dismutase in liver after 1.5 g of diethyldithiocarbamate/kg, showed a maximum decrease (81%) within 1 hour, with a slow return to 64% of normal by 24 hours. Inhibition of superoxide dismutase in vivo and in vitro was confirmed with other assay systems based on the autooxidation of pyrogallol or epinephrine or on reduction of cytochrome c or intro blue tetrazolium. Treatment of animals with diethyldithiocarbamate may provide a useful experimental model to study the role of superoxide dismutase in various tissues.     

Early circulatory and metabolic events in island skin flaps of the pig

Circulatory and metabolic skin-flap events were studied prior to and up to 6 hours after elevation of buttock island flaps in pigs. During the elevation, significant reductions in superficial skin blood flow, measured by laser Doppler flowmetry (LDF) and dermal flap temperature, were seen. Significant correlations were found between blood flow and temperature. Total flap blood flow, measured as venous outflow, also showed an initial transient decrease, but 2 hours after flap construction, venous outflow had returned to preoperative values. A significant increase in lactate release, together with increased oxygen consumption and glucose uptake, was seen 4 hours after the surgical intervention. Hypoxanthine release, indicating ischemia, was seen only during the first hour after flap elevation. Noradrenaline outflow was noted after 4 and 6 hours, but there was no parallel reduction in flap blood flow. A great deal of the flow reduction in acutely elevated island flaps may thus be due to primary hypothermia rather than to the degenerative release of noradrenaline, which seems to have no early effect on skin flap blood flow. On the other hand, the noradrenaline release may be linked to an increased metabolic activity in the skin flaps.     

Pharmacologic action of isoxsuprine in cutaneous and myocutaneous flaps

The therapeutic effects of isoxsuprine on skin capillary blood flow and viability were studied in arterial buttock flaps, latissimus dorsi myocutaneous flaps, and random skin flaps in pigs. It was observed that parenteral isoxsuprine increased capillary blood flow to the skin of arterial buttock flaps and the skin and muscle of latissimus dorsi myocutaneous flaps in a dose-response manner, with a maximum vascular effect observed at 1.0 mg/kg. However, this maximum effective dose of isoxsuprine did not have any significant effect on skin viability in the cutaneous and myocutaneous flaps compared with the control. Examination of the distribution of capillary blood flow within the flaps at varying distances from the pedicle revealed that isoxsuprine did not increase capillary blood flow or perfusion distance in the distal portion of the skin of arterial buttock flaps, latissimus dorsi myocutaneous flaps, and random skin flaps. The increased capillary blood flow as a result of isoxsuprine treatment was limited only to the arterial portion of the arterial buttock flaps and latissimus dorsi flaps. Therefore, it is concluded that isoxsuprine alone is not effective in augmentation of skin viability in cutaneous and myocutaneous flaps. The pharmacologic action of isoxsuprine on the vasculature in the skin and muscle of flaps was also discussed.     

Anti-Inflammatory Effect of Targeted Delivery of SOD to Endothelium: Mechanism,Synergism with NO Donors and Protective Effects In Vitro and In Vivo

Pro-inflammatory activation of vascular endothelium is implicated in pathogenesis of severe conditions including stroke, infarction and sepsis. We have recently reported that superoxide dismutase (SOD) conjugated with antibodies (Ab/SOD) that provide targeted delivery into endothelial endosomes mitigates inflammatory endothelial activation by cytokines and agonists of Toll-like receptors (TLR). The goal of this study was to appraise potential utility and define the mechanism of this effect. Ab/SOD, but not non-targeted SOD injected in mice alleviated endotoxin-induced leukocyte adhesion in the cerebral vasculature and protected brain from ischemia-reperfusion injury. Transfection of endothelial cells with SOD, but not catalase inhibited NFκB signaling and expression of Vascular Cell Adhesion Molecule-1 induced by both cytokines and TLR agonists. These results affirmed that Ab/SOD-quenched superoxide anion produced by endothelial cells in response to proinflammatory agents mediates NFκB activation. Furthermore, Ab/SOD potentiates anti-inflammatory effect of NO donors in endothelial cells in vitro, as well as in the endotoxin-challenged mice. These results demonstrate the central role of intracellular superoxide as a mediator of pro-inflammatory activation of endothelium and support the notion of utility of targeted interception of this signaling pathway for management of acute vascular inflammation.     

Inhibition of ischemia and reflow-induced liver injury by an SOD derivative that circulates bound to albumin

Ischemia followed by reflow often results in tissue injury. Although reactive oxygens seem to play an important role in the pathogenesis of postischemic reflow-induced tissue injury, the mechanism and an efficient way to inhibit oxidative injury are not known. We studied the mechanism by which hepatic transport function was inhibited by a transient occlusion followed by reflow of the portal vein and hepatic artery by using a superoxide dismutase (SOD) derivative (SM-SOD) which circulates bound to albumin with a half-life of 6 h. Occlusion of the hepatic vessels for 20 min followed by reflow for 60 min significantly inhibited transhepatic transport of cholephilic ligands, such as bromosulfophthalein (BSP) and taurocholic acid. Intravenous administration of SM-SOD markedly inhibited the reflow-induced decrease in transhepatic transport of these ligands. Thiobarbituric acid - reactive metabolites (TBAR) in the liver and plasma remained unchanged during occlusion and reflow, while TBAR in the bile increased significantly. Intravenous injection of SM-SOD inhibited the reflow-induced increase in biliary TBAR. Xanthine oxidase activity in plasma also increased during occlusion and reflow by an SM-SOD-inhibitable mechanism. Polymorphonuclear leukocyte-dependent chemiluminescence of the peripheral blood remained unchanged during occlusion, but increased markedly with time after reflow. SM-SOD also inhibited the increase in chemiluminescence almost completely. These and other results suggested that the superoxide radical and/or its metabolite(s) might play an important role in the pathogenesis of the reflow-induced liver injury and that SM-SOD might be useful for studying the mechanism for tissue injury caused by oxygen toxicity.     

Inhibition of posthemorrhagic transfusion-induced gastric injury by a long-acting superoxide dismutase derivative

Although oxygen-free radicals have been postulated to play an important role in the pathogenesis of gastric mucosal injury induced by posthemorrhagic blood transfusion, direct evidence supporting this hypothesis is lacking. Superoxide dismutase (SOD) has been shown to inhibit oxygen toxicity in vitro in various types of cell injury. However, in some cases, oxidative tissue injury cannot be decreased efficiency predominantly due to its rapid elimination by renal glomerular filtration. To overcome such frustrating situations, we have synthesized a SOD derivative that circulates bound to albumin with a half-life of 6 hr. When blood was withdrawn from the rat (22 ml/kg) for 30 min followed by transfusion of the extracted blood, marked gastric mucosal lesions occurred within 30 min after transfusion. Intravenously injected SOD derivative markedly decreased gastric mucosal injury. Kinetic analysis using 125I-labeled albumin revealed that the vascular permeability of the stomach increased significantly after transfusion by a SOD derivative inhibitable mechanism. Thus, superoxide radical and/or its metabolite(s) play a critical role in the pathogenesis of posthemorrhagic transfusion-induced gastric injury.