The effect of ischemia/reperfusion on adenine nucleotide metabolism and xanthine oxidase production in skeletal muscle

Transcription

1 The effect of ischemia/reperfusion on adenine nucleotide metabolism and xanthine oxidase production in skeletal muscle Thomas F. Lindsay, M_D, Shinta Liauw, MSc, Alex D. Romaschin, PhD, and Patti M. Walker, MD, PhD, FRCS(C) Toronto, Ontario, Canada Prolonged ischemia to skeletal muscle as occurs after an acute arterial occlusion results in alterations in adenine nucleotide metabolism. Adenosine triphosphate continues to be used for cellular fimctions, and an ischemia-induced degradation of phosphorylated adenine nucleotides is initiated. In this experiement we demonstrated the time-dependent aspect of adenine nucleotide depletion during ischemia and the production of large quantifies of soluble precursors. In addition, we studied the rate of conversion of xanthine dehydrogenase to xanthine oxidase, a potential source of oxygen-free radicals, after controlled periods of total normothermic ischemia (4 hours and 5 hours) and during the reperfusion phase. During ischemia complete depletion of creatine phosphate occurred in both groups, and adenosine triphosphate fell from to ~mol/gm dry weight after 4 hours and from to I~mol/gm dry weight after 5 hours (p < 0.05). During reperfusion, creatine phosphokinase resynthesis occurred in both groups, but adenosine triphosphate levels were not significantly increased (p > 0.05). A washout of lipid soluble products of adenine nucleotide metabolism occurred equally in both groups. The relationship between phosphorylated adenine nucleotides as measured by the energy charge potential fell significantly in both groups (p < 0.05), but after the shorter period ofischemia (4 hours it returned to normal during early reperfusion but did not after 5 hours of ischemia. There was 21% -+ 4% necrosis after 4 hours and 51% -+ 8% after 5 hours ofischemic stress when assessed at 48 hours. In conclusion, the degree of adenine nucleotide degeneration as determined primarily by the length of the ischemic period, may be the most important determinant of the ultimate extent of skeletal muscle ischemic necrosis that results from an acute interruption of circulation. (J VAse SURG 1990;12:8-15.) Skeletal muscle ischemia, as occurs after an abrupt cessation of circulation, results in an imbalance in the use and production of high energy phosphate molecules. Energy requirements are reduced in ischemic skeletal muscle since no contractions occur, but hydrolysis of adenosine triphosphate (ATP) does continue in order to meet cellular demands for membrane stabilization and ion Compartmentalization. In our previous work 1 we have shown that in skeletal From the Divisions of Vascular Surgery and Clinical Biochemistry, R. Fraser Elliott Vascular Research Laboratory, University of Toronto, Toronto General Hospital. Supported by Medical Research Council of Canada grant MA Reprint requests: Patti M. Walker, MD, Toronto General Hospital, 200 Elizabeth St., EN 9 215, Toronto, Ontario, M5G 2CA. 24/1 / muscle subjected to normothermic ischemia, A'f% ~ levels were maintained at or near preischemic levels for up to 3 hours, predominantly by the depletion of creatine phosphate stores (CP). Subsequently, ATP levels fell linearly until they were less than 20% ofpreischemic values by 6 hours. Parallel studies have shown that 3 hours of ischemic stress causes reversible biochemical and morphologic changes, resulting in little necrosisj After 6 hours of ischemic stress there is complete derangement of cellular structure, and irreversible biochemical alterations occur leading to complete necrosis of the gracilis muscle. After depletion of CP, ATP production is possible only by glycolytic pathways, resulting in the production of lactate, a potentially injurious agent. Restoration of high-energy stores during reperfusion requires an adequate level of precursors, including oxygen and adenine nucleosides, as well as functioning cellular

2 Volume 12 Number 1 Julv 1990 Adenine nucleotide metabolism 9 mechanisms for electron transport in the mitochondria. 3 Ischemic damage may prevent normal cellular synthetic processes, and ischemic degeneration ofnucleotides can produce lipid soluble products that are washed out of the cytoplasm on reperfusion. In addition, a time-dependent mechanism for increasing necrosis with increasing periods of ischemia has been postulated by McCord et al.4 They suggested that the breakdown of adenine nucleotides resulted in increased production of hypoxanthine and xanthine, and that the increased production coupled with ischemia-induced conversion of xanthine dehydrogenase to xanthine oxidase results in free radical damage on reintroduction of molecular oxygen on reperfusion, the extent of which is dependent on the quantity of precursors available. To extend the understanding of this derangement in adenine nucleotide metabolism, particularly with regard to precursor washout during reperfusion, we have studied the effects of 4 and 5 hours of normothermic ischemia followed by reperfusion in the canine gracilis muscle model. Using high performance liquid chromatography we analyzed ATP, adenosine diphosphate (ADP), and adenosine monophosphate (AMP), inosine monophosphate, adenosine, inosine, hypoxanthine, xanthine, and urate. Along with the absolute levels of these metabolites, we determined the relationship between concentrations of the high energy phosphate nucleotides using the energy charge potential (ECP) as suggested by Atkinson. 5 We characterized the conversion of xanthine dehydrogenase to xanthine oxidase during prolonged ischemia and during in vivo reperfusion as a putative source of oxygen free radicals. METHODS Model The bilateral canine gracilis muscle model was used. 6 Both gracilis muscles were dissected free of all collateral circulation, and the minor pedicle was ligated. Both tendons were divided and then resutured to maintain resting muscle length. The arterial inflow and venous outflow were isolated to the major pedicle. Periods of ischemia were induced by the application of microvascular clips to both the artery and vein of the major pedicte. Biopsy specimens (50 mg wet weight) were taken before the application of the microvascular clips, at the end of the ischemic period, and at 5, 15, 30, and 45 minutes ofreperfusion. After the biopsies the skin was oversewn, and the animal survived. Forty-eight hours after the experiment the animal was killed, and both gracilis muscles were harvested. They were sliced into six slices, and the extent of necrosis was determined by tetrazolium staining and computerized planimetry, as we have described elsewhere. The skeletal muscle biopsies were freeze-dried overnight, and 5 to 10 mg of the dry muscle fibers were separated from cormective tissue. Tissues were extracted with ice cold 0.5 mol/l perchloric acid with addition of internal standard 2-0- methytadenosine. The supernatant was collected and neutralized with 2 mol/l potassium hydroxide by use of a mixture of indicators, phenol red and bromothymol blue (each 0.1%, mixed 1:1). The mixture was back titrated to ph 6 by use of 0.1 mol/l perchloric acid. The neutralized supernatant was freezedried and stored at -80 C until analysis. Samples were analyzed for purinc nucleotides on a Waters High Performance Liquid Chromatography system (Waters Chromatography Div., Millipore, Milford, Mass.), which included dual pumps (model 501) controlled by a solvent programmer (model 660), a programmable multiwavelength detector (model 490), and dual data modules (model 740). A 5 ~m Radial Pak Resolved C-18 column was used as the stationary phase (Waters Chromatography Div). 8 The mobitc phase solvents were 0.1 mol/l ammonium phosphate buffer ph 5.7 (solvent A) followed by 40% methanol in water (solvent B). Each sample was reconstituted in solvent A and injected. The flow rate was set at 1.5 mi/min, and the solvent sequence was 15 minutes of solvent A followed by an exponential gradient of solvent B (number 8 on model 660 solvent programmer) over 5 minutes, then 100% solvent B for the final 10 minutes. The equilibration time between injections was 10 minutes with solvent A. The nucleotide clution was followed at 254 um for a 30-minute period. The areas under each peak were integrated by the data module. Purine standards were prepared to map retention times and evaluate adequacy of separation. Five standard concentrations were chromatographed after addition of internal standard for each purine to be quantified to determine individual slope factors (K) for each analyte. To obtain the concentration of the nucleotide in the injected sample the following formula was used: C(N) = K* C(OMA)/A(OMA) / dil(oma) where K was the slope factor for each nucleotide, A(N) and A(OMA) were the area of the individual nucleotide and internal standard (2-0- methyladenosine = OMA) from the data module

3 10 Lindsay et al. Journal of VASCULAR SURGEK,Y Table I. The levels of tissue adenine nucleotide and purine bases after 4 hours and at PI (preischemia), EI (end ischemia), and at 5, 15, 30, 45 minutes of reperfusion. (mean + SEM, n = 8) Time ATP ADP AMP ADO IMP INO HYPOX XAN UA NAD TN CP PI EI m in m in m in m in ATP, Adenosine triphosphate; ADP, adenosine diphosphate; AMP, adenosine monophosphate; ADO, adenosine; IMP, inosine monophosphate; INO, inosine; HYPOX, hypozanthine; XAN, xanthine; UA, urate; TN, total nucleotides; CP, creatine phosphate, NAD, nicotinamide adenine dinucleotide. Table II. The levels of tissue adenine nucleotide and purine bases after 5 hours of ischemia and at PI (preischemia), EI (end ischemia), and at 5, 15, 30, 45 minutes of reperfusion. (mean + SEM, n= 8) Time ATP ADP AMP ADO IMP INO HYPOX XAN UA NAD TN CP PI EI m in m in m in m in ATP, Adenosine triphosphate; ADP, adenosine diphosphate; AMP, adenosine monophosphate; ADO, adenosine; IMP, inosine monog phosphate; INO, inosine; HYPOX, hypoxanthine; XAN, xanthine; UA, urate; TN, total nucleotides; CP, creatine phosphate, NAIl: nicotinamide adenine dinucleotide. output, C(OMA) is concentrataon of OMA in the injected sample. The final concentrations were determined after correcting for the weight of tissue extracted and the initial volume. Concentrations were expressed in micromoles per gram dry weight of tissue; XANTHINE OXIDASE / DEI-IYDROGENASE ENZYME ANALYSIS The enzymes xanthine dehydrogenase and oxidase were prepared from skeletal muscle according to the method of Roy and McCord. 9 Tissues freezeclamped in liquid nitrogen were diluted 1 : 5 with 0.1 mol/l potassium phosphate buffer (ph 8.1) conraining 1 mmol/l phenylmethylsulfonylfuoride and 10 mmol/l dithiothreitol (buffer A). Tissues were homogenized with a Polytron homogenizer (Brinkmann Instruments, Inc., Westbury, N.Y.) at a speed of 7 for three consecutive 30-second bursts. The tissue homogenate was kept in an ice bath at 4 C throughout the homogenization. The homogenate was subsequently centrifuged at 27,000g for 30 minutes, the upper lipid layer was aspirated off, and the remaining supematant was decanted. This supernatant was then chromatographed on a Sephadex (Pharmacia, Uppsala, Sweden) G-25 column (1 20 cm) by applying 2 ml portions of the supematant on the column and eluting with buffer A. One milliliter fractions were collected and assayed for enzyme activity. The spectrophotometric method of Stirpe

4 Volume 12 Number 1 Juh~ 1990 Adenine nucleotide metabolism ll - "o E O E a_ o : 8o! o 20 ~ 4 hr 10 0 I I I I I I I I I I PI El Time (minutes) Fig. 1. Creatine phosphate (CP) levels in p~mol/gm/dry weight, at preischemia (PI), end ischemia (EI), and at 5 to 45 minutes of reperfusion. Solid dots = 4 hours and open dots = 5 hours of ischemic stress (mean + SEM, n = 8). and Della Corte ~ was used to assay xanthine dehydrogenase and oxidase in canine intestine and liver and rat plus tenusismus muscle. Enzyme assays were performed at 37 C by measuring the production of urate at 292 nm spectrophotometrically by use of a temperature controlled Beckman (Beckman Instruments, Inc., Fullerton, Calif.) spectrophotometer (model DU-40) or a Cobas Fara Centrifugal Analyzer (Roche Diagnostics, Brantford, Ontario). The oxidase assay mixture c,.,ntained 1 ml of 0.1 mol/l potassium phosphate buffer at ph 8.1, 50 mm 3 of 2 mmol/l xanthine, and 50 to 200 mm 3 of the enzyme preparation. In the assay that measured both the dehydrogenase and oxidase forms of the enzyme, 50 rrun 3 of 17 mmol/l NAD + was added. Rates of urate production were calculated from a standard curve of absorbance at 292 run plotted against increasing concentrations of urate. To ensure that xanthine dehydrogenase or oxidase was responsible for the rates measured, allopurinol (10 mm 3 of 2 ~zmol/l solution) was added to assays of both D and O forms. The enzyme assay conditions were standardized, and the sensitivity was determined by use of purchased xanthine oxidase (EC ) (Sigma Co., St. Louis, Mo.). Serial dilutions of the enzyme were made with buffer A, and the activity was measured in duplicate. The sensitivity was defined as the largest dilution that maintained a linear relationship between absorbance and dilution factor. To ensure that the enzyme purification and tissue assay procedure was satisfactory, xanthine dehydrogenase and xanthine oxidase were partially purified and assayed in canine liver and small intestine where high concentrations have been reported, n Protein was detcrmined with thc comassie blue method of Bradford.12 A sensitive spectrofluorometric assay was used to measure xanthine dehydrogenase/oxidase in canine skeletal muscle? 3 The assay used pterin as substrate to measure the oxidase reaction and pterin plus methylene blue to measure both oxidase and dehydrogenase. Allopurinol (20 txmol/l) was added to inhibit the enzymes and conform the specificity of measurement. Isoxanthopterin was used to standardize enzyme activity by constructing a plot of fluorescence versus concentration. Fluorescence was monitored in a Perkin Elmer (Perkin-Elmer Corp., Norwalk, Conn.) S spectrofluorometer by exiting at 345 nm and measuring emission at 390 nm. RESULTS The levels (mean + SEM, n = 8) of tissue adenine nucleotides, nucleosides, and the corresponding bases for preischemia, end ischemia, and at 5, 15, 30, and 45 minutes ofreperfusion are shown in Table I after 4 hours of ischemia and in Table II after 5 hours of ischemia. The total tissue content of all

5 12 Lindsay et al. Journal of VASCULAR SURGE_P~Y 1.0 o--- 5 hr == 4 hr 0.9 ~ UJ , i i i I i t a i i PI El Time (minutes) Fig. 2. Energy charge potential (ECP) at preischemia (PI), end ischemia (EI), and at 5 to 45 minutes of reperfusion. Solid dots = 4 hours and open dots = 5 hours of ischemic stress (mean + SEM, n = 8). adenine nucleotides and breakdown products are equal at preischemic levels and postischemic levels for both groups, but the relative concentrations varied dramatically. Adenosine triphosphate concentration had fallen to 3.9 _ 0.8 ~mol/gm dry weight after 5 hours of ischemia compared to 4 hours of ischemia where the level had decreased to ~mol/gm dry weight (p < 0.05). However, during the reperfusion phase there was no significant increase in ATP stores in either groups (p > 0.05, ANOVA). There was a significant difference in the measured ADP levels after ischemia ,6 ~mol/grn dry weight after 4 hours and ~mol/grn dry weight after 5 hours (p < 0.05). There was a very significant accumulation of the precursors of adenine nucleotides as demonstrated by the increased levels ofinosine and hypoxanthine, after both 4 and 5 hours of ischemic stress. Creatine phosphate levels had fallen to equally low levels after 4 and 5 hours of ischemia (greater than 90% depleted). The behavior during the reperfusion, however, was significantly different between the two groups (Fig. 1). Immediately on reperfusion after 4 hours of ischemia, CP levels increased to ~mol/grn dry weight or near preischemic levels. After 5 hours of ischemia the level had risen to ~mol/mg dry weight during early reperfusion (p < 0.05). There was also a significant overshoot of CP stores during reperfusion after a shorter period of ischemia 4 hours, compared to the 5-hour group, ( ) versus ( ) (p < 0.05), respectively. In Fig. 3, the relationship between the ATP/ADP ratio and intraceilular levels of CP is shown. This linear relation is significant for both 4 and 5 hours of ischemia (r = 0.83, p < 0.05), indicating an appropriate response in the mitochondria for production of highenergy phosphate bonds. There was also a statistically significant depletion in total adenine nucleotides and their breakdown products, from the similar levels preischemia and immediately end ischemia, and at 45 minutes of repe~ fusion from to 21.8 _-_ 2.7 after 4 hours ofischemia and from to after 5 hours of ischemia, p < Atkinson 5 has suggested that the ECP may better reflect the availability of high energy stores for the synthetic process rather than simply the absolute levels. As demonstrated in Fig. 2, ECP felt to a significantly lower level after 5 hours of ischemia compared to 4 hours ofischemia (p < 0.05). In addition, during reperfusion after 4 hours ofischemia the ECP had returned to normal in 5 minutes and remained at that level throughout the period of observation. In the 5-hour group, however, destined for more significant necrosis, the ECP fell to a significantly lower level and remained low throughout the entire period of reperfusion. In Fig. 3, the relationship between the ATP/ADP ratio and intramuscular levels of CP is demonstrated. Their linear relation is sta-

6 Volume 12 Number 1 It'"', 1990 Adenine nucieotide metabolism 13 O,. E3 12. I.- 3 o 5hr 4h 0 i, I I I, i I CP (g moles/gr dry weight) Fig. 3. The relationship between ATP/ADP ratio and intracellular stores of CP is demonstrated at end ischemia and at 5, 15, 30, 45 minutes of reperfusion (n = 8, r = 0.72, p < 0.05). tistically significant for both 4-hour and 5-hour experiments (r = 0.83, p < 0.005). The final extent of necrosis was 21% _+ 4% after 4 hours ofischemia and 51% _+ 8% after 5 hours of ischemia (n = 8, + SEM). The assay ofxanthine dchydrogenasc and oxidase in canine tissue including liver, bowel, and skeletal muscle confirmed that the activity of dehydrogenase and oxidase enzymes in the bowel and liver to be much greater than in skeletal muscle (Table III). In bowel the dehydrogenase activity rapidly converted ~o oxidase during incubation under ischcmic conditions with levels of total activity of units/gm of protein demonstrated with 40% conversion to xanthine oxidase. No activity of xanthine dehydrogenase and xanthine oxidase was detected in the canine gracilis muscle biopsy specimen taken after ischemia and through reperfusion by use of the spectrophotometric assay. Using the more sensitive spectrofluorometric assay, a level of < unit/gm of protein of total enzyme activity was present in skeletal muscle with 25% conversion to xanthine oxidase after prolonged ischemia and reperfusion. DISCUSSION The results of this study demonstrate a timedependent breakdown of high-energy stores during extended periods of ischemia and only partial recovery during early reperfusion. Both 4-and 5-hour De- Table III. Mean XD + XO activity + SD Units/gin Tissue protein Mean % XO Canine *2,5 --_ 0.6 <1 Small intestine (after ~-2.3 +_ rain ischemia) Canine fiver * , <1 Canine gracilis muscle t _ (nonischemic) Canine gracilis muscle t (after 5 hours ischemia) XD, Xanthine dehydrogenase; XO, xanthine oxidase. *Spectrophotometric assay. tspectrofluorometric assay. riods of ischcmia completely depleted CP levels, and 5 hours resulted in ATP stores being depleted by 80%, and 4 hours resulted in ATP stores being depleted by 50%. The accumulation of lipid soluble precursors including adenosine, inosinc, hypoxanthine, and xanthincs during ischemia demonstrates significant activity of the enzymes 5'nucleotidase, AMP deaminase, adenosine deaminase, and low levels of activity of xanthine dehydrogenasc/oxidasc in skeletal muscle. The gradual decline during repcrfusion in the tissue concentrations of these precursors capable of crossing cell membranes is in keeping with our previous data on the degree of reactive hyperemia that occurs after a period ofischemia, in that the peak

7 14 Lindsay et al. Journal of VASCULAR SURGE~ period of increased blood flow is not immediately present on reperfusion, but rather is 15 to 30 minutes later. The rate of the washout of inosine agrees specifically with both the delay in peak blood flow as well as the fact that peak flow is not as great after 5 hours of ischemia as it is after 4 hours. ~4 The failure of resynthesis of ATP during reperfusion agrees with reports suggesting that de novo synthesis of these phosphorylated nucleotides is a slow process and is unlikely to contribute to cell survival by increased energy availability, and production through salvage pathways after such an ischemic period is limited, is Morphologic information can be projected from the overshoot of CP production during early reperfusion. Since CP acts as a shuttle in the mitochondrial membrane, carrying high energy phosphate bonds to convert ADP to ATP in the cytoplasm, 16 the relative increase in CP suggests a functioning mitochondrial oxidative rephosphorylation system and a relative lack of cytoplasmic ADP to act as an acceptor.17 These factors suggest that mitochondria may well still be viable after extended periods of ischemia, ~8 and that other factors are responsible for the failure to restore energy stores. The relationship between phosphorylated adenine nucleotides, as demonstrated by the ECP, suggest that after a shorter period of ischemia (4 hours), even though ATP is not returned to preischemic levels, there is an ability to maintain phosphorylated nucleotides in a higher state of energy. In comparison, after 5 hours of ischemia, which results in 51% necrosis, the ECP remains reduced through the entire period of observed reperfusion. These findings are in keeping with studies on ischemia/reperfusion injury to the myocardium where a lack of precursors, rather than mitochondrial dysfunction, has been suggested as a major determinant of lack of restoration of ATP. 19 Increased rate of replenishment has been demonstrated by the addition of adenosine, inosine, and 5-amino imidazole-4-carboxamide riboside (AICAR). However, despite increased levels of ATP, improvement in myocardial contractility has not been noted. This dichotomy between restoration of high-energy phosphate stores and lack of improvement in functional parameters, however, is not inconsistent with our data. Energy stores required for viability may differ greatly compared with those required for contractile activity, and in addition, cytoplasmic organelles are required to allow myofibril shortening, (for example, the sarcoplasmic reticulum), which may themselves suffer ischemia/reperfusion injury rendering them nonfunctional. The reported role of xanthine dehydroge- nase / xanthine oxidase conversion on free-radical mediated injury in skeletal muscle may be less significant than has been suggested. 2 This mechanism of injury is an attractive hypothesis, as its importance has been demonstrated in the intestine, 2~ and safe pharmacologic inhibition of this enzyme exists (oxypurinol) allowing potential clinical interventions. However, these studies have demonstrated that despite the apparent abundance of precursors (xanthine and hypoxanthine) there is very little increase in the product of the reaction catalyzed by the enzyme, namely uric acid. Despite relatively sensitive methods for analysis, only small amounts of xanthine dehydrogenase were detected, and conversion to xanthine oxidase was very slow despite a prolonged ischemic insult and adequate sampling during the reperfusion phase: This is in direct contrast with the small bowel, wher~ large amounts of xanthine dehydrogenase are present, and ischernia-induced conversions to the oxidase occur rapidly. A species variability in xanthine dehydrogenase/xanthine oxidase levels has been demonstrated previously, but the importance of this mechanism of injury may be insignificant in human skeletal muscle, in keeping with recent evidence that suggests that human myocardial tissue contains little activity of this enzyme system. 22 The source of oxygen-free radical production may result from activated leukocytes; however, their role in skeletal muscle has yet to be evaluated. In conclusion, there is a time-dependent, ischemia-induced breakdown of phosphorylated adenine nucleotides in skeletal muscle with a buildup of lipid soluble precursors. During reperfusion there is a washout of these precursors and failure of resyrf~ thesis of ATP stores. The relationship between phos ~= phorylated adenine nucleotides (ATP, ADP, AMP) as shown by the Atkinson ECP is reduced during 4 hours of ischemia but recovers quickly on reperfusion, and only minimal necrosis results. After 5 hours of ischemia, ATP falls further, the ECP does not recover during reperfusion, and a significantly greater amount of ischemic necrosis results. The overshoot of CP production during reperfusion suggests functioning mitochondrial oxidative rephosphorylation pathways. Despite an abundance of precursors, xanthine oxidase/xanthine dehydrogenase levels remain low during reperfusion, and it is unlikely to be a significant mechanism of injury. These factors suggest that the degree of energy depletion that occurs during ischemia may be the single most important determinant of the ultimate extent of ischemic damage that follows an acute arterial occlusion. Although enhancement of ischemic injury by events during re-

8 Volume 12 Number 1 Jr' Adenine nudeotide metabolism 15 perfusion can be influenced, a limit on the effectiveness of these therapeutic interventions may well exist. Clinically, identifying this limiting factor may be important to avoid attempted salvage of muscle already destined not to be viable, and thus exposing patients unnecessarily to the known metabolic and hemodynamic sequelae of reperfused dead muscle. Currently the duration of the ischemic interval and the degree of ischemia present are the best clinical clues of nonviability. Further investigation may define more precise criteria ofnonviability thus avoiding the sequelae caused by reperfusion of dead muscle. REFERENCES 1. Harris K, Walker P, Mickle D, et al. Metabolic response of skeletal muscle to ischemia. Am J Physiol 1986;250:H Labbe R, Lindsay T, Walker P. The extent and distribution of skeletal muscle necrosis after graded periods of complete ischemia. J VAsc SURG 1987;9: Hanzilkova V, Gutmann E. Effect of ischemia on contractile and histochemical properties of rat soleus muscle. Pfleugers Arch., 1979;279: McCord J, Oxygen derived free radicals in post ischemic tissue injury. N Engl J Med 1985;312: Atkinson D. The energy charge of the adenylate pool as a regulatory parameter interaction with feedback modifiers. Biochemistry 1966;5: Kuzon W, Walker P, Mickle D, Harris K, Pynn B, McKee N. An isolated skeletal muscle model suitable of acute ischemic studies. J Surg Res 1986;41: Labbe R, Gadey R, Lindsay T, et al. Quantitation of post ischemic skeletal muscle necrosis: histochemicala and radioisotope techniques. J Surg Res 1987;43: Hull-Ryde E, Lewis W, Veronee J, Lowe J. Simple step gradient elution of the major high energy compounds and their catabolites in cardiac muscle using high performance liquid chromatography. J Chromatog 1986;377: ). Roy R, McCord J. Superoxide and ischemia: conversion to xanthine dehydrogenase to xanthine oxidase. In: Proceedings of the Third International Conference on Superoxide and Superoxide Dismutase. Greenwald R~ Cohen G, eds. New York: Elsevier/North Holland, 1983: Stirpe F, Della Corte E. The regulation of rat liver xanthine oxidase: conversion of type D into type O by thermolabile factor and reversibility by dithioerythritol. Biochem Biophys Acta 1970;212: Granger D, Rntili G, McCord J. Superoxide radicals in feline intestinal ischemia. Gastroenterology 1981;81: Bradford M. A simple and sensitive assay of protein based on Coomassie Blue G-250 dye binding. Anal Biochem 1976; 72: Marldey H, Faillace L, Mezey E. Xanthine oxidase activity in brain. Biochem Biophys Acta 1973;309: Forrest I, Lindsay T, Romaschin A, Mickle D, Walker P. The rate and distribution of blood flow following prolonged skeletal muscle ischemia. J VASC SURG 1989;10: Tullson P, John-Adler H, Hood D, Terjtmg R. De novo synthesis of adenine nucleotides in different skeleral muscle fiber types. Am Physiot Soc 1988;C Meyer R, Sweeney A, Kushmerick M. A simple analysis of the phosphocreatine shuttle. Am J Physiol 1984;246:C Jacobus W. Respiratory control and the integration of heart high energy phosphate metabolism by mitochondrial creatine kinase. Ann Rev Physiol 1985;47: Humphrey S, Hollis D, Cartner L. The influence ofinhibitors of the ATP degradative pathway on recovery of function and high energy phosphate after transient ischemia in the rat heart. J Mol Cell Cardiol., 1986;18([Suppl 4]): Ambrosio G, Jacobus W, Mitchell M, Litt M, Becker L. Effects of ATP precursors in ATP and free ADP content and functional recovery of post ischemic hearts. Am J Physiol 1989;256:H Korthius R, Granger D, Townsley M, Taylor A. The role of oxygen derived free radicals in ischemia induced increases in canine skeletal muscle vascular permeability. Circulation 1985;72: Parks D, Granger N. Xanthine oxidase: biochemistry, distribution and physiology. Acta Physiol Scand Suppl 1986; 548: Eddy L, Stewart J, Jones S, Engerson T, McCord J, Downey J. Free radical producing enzyme xanthine oxidase is undetectable in human hearts. Am J Physiol 1987;253:H