Evidence that mitochondrial respiration is a source of potentially toxic oxygen free radicals in...

THE JOURNAL OF B ~ L O G I C A L CHEMISTRY 0 1993 by The Amencan Society for Biochemistry and Molecular Biology, Inc. Vol. 268, No. 25, Iesue of September 5, pp. 18532-1@541,1993

Printed in U. S. A.

Evidence That Mitochondrial Respiration Is a Source of Potentially Toxic Oxygen Free Radicals in Intact Rabbit Hearts Subjected to Ischemia and Reflow’

(Received for publication, December 17, 1992, and in revised form, March 31, 1993)

Giuseppe AmbrosioS, Jay L. ZweierjT, Carlo Duilio, Periannan Kuppusamyj, Giuseppe Santoro, Pietro P. Elia, Isabella Tritto, Plinio Cirillo, Mario Condorelli, Massimo Chiariello, and John T. Flahertyj From the Department of Medicine, Division of Cardiology, Second School of Medicine, University of Naples, 80131 Naples, Italy and the $Department of Medicine, Division of Cardiology, Johns Hopkins University School of Medicine, Baltimore, Maryland 21205

Previous in vitro studies have shown that isolated mitochondria can generate oxygen radicals. However, whether a similar phenomenon can also occur in intact organs is unknown. In the present study, we tested the hypothesis that resumption of mitochondrial respiration upon reperfusion might be a mechanism of oxygen radical formation in postischemic hearts, and that treatment with inhibitors of mitochondrial respiration might prevent this phenomenon. Three groups of Lan- gendorff-perfused rabbit hearts were subjected to 30 min of global ischemia at 37 “C, followed by reflow. Throughout ischemia and early reperfusion the hearts received, respectively: (a) 6 mM KC1 (controls), (b) 6 mM sodium amobarbital (Amytal“, which blocks mitochondrial respiration at Site I, at the level of NADH dehydrogenase), and (c) 6 mM potassium cyanide (to block mitochondrial respiration distally, at the level of cytochrome c oxidase). The hearts were then processed to directly evaluate oxygen radical generation by electron paramagnetic resonance spectroscopy, or to measure oxygen radical-induced membrane lipid peroxidation by malonyl dialdehyde (MDA) content of subcellular fractions. Severity of ischemia, as assessed by “P-nuclear magnetic resonance measurements of cardiac ATP, phosphocreatine, and pH, was similar in all groups. Oxygen-centered free radical concentration averaged 3.84 f 0.64 PM in reperfused control hearts, and it was significantly reduced by Amytal treatment (1.98 2 0.26; p < 0.06), but not by KCN (2.68 f 0.96 PM; p = not significant (NS)), consistent with oxygen radicals being formed in the mitochondrial respiratory chain at Site I. Membrane lipid peroxidation of reperfused hearts was also reduced by treatment with Amy- tal, but not with KCN. MDA content of the mitochondrial fraction averaged 0.76 f 0.06 nM/mg protein in controls, 0.72 f 0.06 in KCN-treated hearts, and 0.64

* This work has been supported in part by Grant 91.00122.PF41 from Consiglio Nazionale delle Ricerche (Progetto Finalizzato Prev- enzione e Controllo dei Fattori di Malattia), by Specialized Center of Research in Ischemic Heart Disease Grant P50HL 17655, by Grant HL-38324 from the National Institutes of Health, and by NATO International Research Grant 0139/88. A preliminary account was presented at the 65th Meeting of the American Heart Association, November 16-19,1992, New Orleans, LA. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

$ To whom correspondence should be addressed Cattedra di Car- diologia, I1 Facolta’ di Medicina, Via S. Pansini 5, 80131 Naples, Italy. Tel.: 39-81-7462216; Fax: 39-81-7462229.

ll Established Investigator of the American Heart Association.

f 0.06 in Amytal-treated hearts ( p < 0.06 versus both groups). Similarly, MDA content of lysosomal membrane fraction was 0.64 f 0.09 nM/mg protein in controls, 0.79 C 0.16 in KCN-treated hearts, and 0.43 2 0.06 in Amytal-treated hearts ( p 0.06 versus both groups). Since the effects of Amytal are known to be reversible, in a second series of experiments we investigated whether transient mitochondrial inhibition during the initial 10 min of reperfusion was also associated with beneficial effects on subsequent recovery of cardiac function after wash-out of the drug. At the end of the experiment, recovery of left ventricular end- diastolic and of developed pressure was significantly greater in those hearts that had been treated with Amytal during ischemia and early reflow, as compared to untreated hearts. In conclusion, our data demonstrate that in intact hearts electron flow through the respiratory chain may be an important source of oxygen radicals, which may form at the sites of interac- tions between Fe-S clusters and ubiquinone, and that resumption of mitochondrial respiration upon reoxygenation might contribute to reperfusion injury.

Restoration of flow after a period of ischemia is accompanied by generation of large amounts of potentially toxic oxygen free radicals (1-4), and it has been proposed that this phenomenon may account for the occurrence of a specific form of reperfusion-mediated tissue damage (5, 6). In reperfused hearts, oxygen radicals can be generated by several mechanisms, including the xanthinelxanthine oxidase reaction (7, 8), and the activity of NADPH oxidase and myelo- peroxidase of activated leukocytes, which migrate into the previously ischemic area (9). Another potential source of oxygen radicals is thought to be mitochondrial respiration (10,l l) . In the process of mitochondrial electron transport, oxygen

is normally reduced to water through several steps in which hydrogen atoms act as electron donors (Fig. 1). However, studies on isolated mitochondria have shown that oxygen can also undergo 1-electron reduction, with formation of superoxide radicals (‘0;) and hydrogen peroxide (Hz02) (10-17). At least two sites in the mitochondrial respiratory chain have been identified where oxygen radicals may be generated NADH dehydrogenase (15, 16) and ubiquinone (17) (Fig. 1). At both sites, ’ 0; formation appears to derive from 1-electron transfer through iron-sulfur clusters, with formation of semiquinone, which eventually oxidizes yielding ’ 0; and then

18532

Mitochondria and Reperfuswn Injury 18533

H&. It is estimated that under normal conditions 1-2% of oxygen utilized by mitochondria leads to the formation of superoxide radicals (11). This "physiologic" generation of oxygen radicals is normally inactivated by endogenous scavenger mechanisms present within the cells (5). However, data derived from i n vitro experiments suggest that oxygen radical generation can be greatly enhanced when mitochondrial respiration is stimulated under conditions of altered redox state and of decreased availability of ADP (10-17). Such conditions are likely to occur in hearts reperfused after ischemia. For- mation of oxygen radicals might, in fact, drastically increase upon resumption of oxidative phosphorylation at the time of reflow, when oxygen is made again available to mitochondria that have accumulated large amounts of reducing equivalents, while decreasing their adenine nucleotide content, during the period of ischemia. In addition, decreased activity of oxygen radical scavenging enzymes may further aggravate the imbal- ance between generation and detoxification of oxygen radicals in postischemic hearts (18-20). It has therefore been proposed that mitochondrial respiration might be an important source of oxygen radicals and, hence, a potential contributor to reperfusion injury (5). This hypothesis is indirectly supported by the observation that a prominent generation of oxygen radicals could be induced in vitro upon reoxygenation of mitochondria isolated from hearts that had been subjected to ischemia (21). However, no data are currently available to demonstrate that this phenomenon can actually occur i n intact organs undergoing ischemia and reflow.

The present study was therefore designed to test whether resumption of mitochondrial oxidative phosphorylation upon postischemic reflow can be a source of toxic oxygen radicals in isolated rabbit hearts. This experimental model has been previously characterized with respect to various aspects of oxygen radical toxicity (1, 4, 22-25). In addition, it has the distinct advantage of allowing assessment of the contribution of mitochondrial respiration to oxygen radical generation without the confounding effects of two other major sources of free radicals, namely xanthine oxidase activity, which is very low or absent in rabbit hearts (26,27), and leukocytes, which are absent from the perfusion buffer. Oxygen radical formation was evaluated by electron paramagnetic resonance spectroscopy in untreated hearts subjected to ischemia and reflow and in hearts in which mitochondrial respiration during early reflow was blocked by administration of either sodium amobarbital (Amytal"', an inhibitor of NADH dehydrogenase; Ref. 28), or cyanide (which blocks electron flow at the level of cytochrome c oxidase; Ref. 28) (Fig. 1). Since lipid peroxidation is a major biochemical consequence of oxygen radical attack (25, 29, 30), we also investigated whether attempts to inhibit mitochondrial generation of oxygen radicals were accompanied by a concomitant effect on membrane lipid peroxidation of reperfused hearts. Finally, in another series of experiments we investigated whether transient inhibition of mitochondrial respiration during the early minutes of reflow would also reduce the development of reperfusion injury, as assessed by recovery of contractile function in postischemic hearts.

MATERIALS AND METHODS

Isolated Heart Preparation Female New Zealand White rabbits (1.5-2.0 kg) were heparinized

and anesthetized with intraperitoneal pentobarbital. The hearts were removed, the ascending aorta was cannulated, and retrograde perfusion was started under constant pressure (80 mm Hg) with perfusate containing 117 mM sodium chloride, 6.0 mM potassium chloride, 3.0 mM calcium chloride, 1.0 mM magnesium sulfate, 0.5 mM EDTA, 16.7 mM glucose, and 24 mM sodium bicarbonate, pH 7.4. The perfusate

was equilibrated at 37 "C with a 95/5% mixture of 0 2 and COS, and not recirculated. The hearts were paced at 175 beats/min by means of a wick electrode (containing saturated KCl) connected to a Grass SD-9 stimulator. To assess contractile function, a latex balloon, connected to a Statham P23Db transducer, was inserted into the left ventricular cavity through the mitral opening and secured with a ligature that included the left atrial remnants. The balloon was initially inflated with saline to produce an end-diastolic pressure of 10 mm Hg, which is on the plateau of the end-diastolic volume/end- systolic pressure curve for this preparation. Coronary flow was measured by aspiration of perfusate overflow.

Experimental Protocol After equilibration, the hearts were subjected to 30 min of global

ischemia at 37 "C, followed by reperfusion. At the time of ischemia, the hearts were divided into three groups which received, respectively: (i) 5 mM KC1 (control hearts); (ii) 5 mM Amytal (to block mitochondrial NADH dehydrogenase; Ref. 281, plus 5 mM KCk (iii) 5 mM potassium cyanide (to block mitochondrial cytochrome c oxidase; Ref. 28).

Five mM KC1 was added to groups I and I1 hearts, to control for the additional 5 mmol/liter K+ ions given to KCN-treated hearts (group 111). Global ischemia was induced by interrupting the aortic inflow. The drugs were administered by intracoronary infusion during the initial 5 min of ischemia via a side arm in the perfusion line, at the rate of 0.5 ml/min, which corresponded to <2% of base-line coronary flow. During this early phase of ischemia, the KCl, the Amytal, and the KCN were dissolved in buffer equilibrated with 95% Nz and 5% COz, to prevent residual oxidative phosphorylation which might have otherwise occurred because of perfusion with oxygen- containing drug solutions. Amytal and KCN solutions were brought to pH 7.4-7.6 by adding a small amount of diluted HC1 and were used immediately. Drug infusion was then stopped, and the hearts were subjected to an additional 25 min of zero-flow ischemia. During the ischemic period the hearts were maintained at 37 "C by a continuous flow of warm perfusate around the heart. At the onset of ischemia, the balloon was deflated and the hearts allowed to beat spontaneously. The three groups were reperfused with buffer containing 5 mmol/ liter of the same test drug infused during ischemia. Three sets of experiments were performed.

EPR Experiments-To assess the effects of inhibition of mitochondrial respiration on myocardial generation of free radicals, ischemic hearts (n = 7 for each of the three groups) were reperfused for 30 s (which corresponds to the time of peak free radical generation in this model; Refs. 1,4, and 24), and then immediately processed for electron paramagnetic resonance (EPR) measurement of free radical concentrations (see below).

NMR and Lipid Peroridation Experiments-Hearts (n = 6 for each of the three groups) were subjected to ischemia, reperfused for 10 min, and then processed to measure lipid peroxidation of subcellular membranes, as a biochemical marker of oxygen radical attack (see below). To document that the severity of ischemia was similar in controls and in treated hearts, changes in tissue concentration of high energy phosphate metabolites and intracellular pH were monitored using 31P NMR spectroscopy (see below).

Oxygen Consumption Experiments-To document that mitochondrial respiration was blocked by the two treatments, O2 consumption was measured in three additional hearts for each group subjected to ischemia and reperfusion. At the end of the 10-min infusion of the various test drugs, reperfusion was continued in all hearts with standard perfusate for an additional 10 min (wash-out). Oxygen consumption was measured at base-line level and after 1, 5, 10, and 20 min of reperfusion (see below).

Electron Pararnugnetic Resonance Experiments Direct EPR spectroscopy of frozen tissue was employed to detect

the radical signals arising from reperfused hearts. This methodology may be better suited than spin-trap technique to detect intracellularly formed free radicals. In particular, only direct EPR spectroscopy of frozen tissues can detect changes in redox state of components of the mitochondrial electron transport chain. Care was taken to minimize artifactual generation of radical signals, as detailed below.

Just before ischemia, the intraventricular balloon and pacing electrode were removed. After 30 s of reflow, the hearts were freeze- clamped using Wollenberg tongs precooled in liquid nitrogen. The freeze-clamped hearts were briefly ( 4 min) and coarsely (particle diameter > 2 mm) ground under liquid nitrogen in a ceramic mortar,

18534 Mitochondria and Reperfusion Injury

FIG. 1. Schematic representation of the mitochondrial respiratory chain, in which molecular oxygen normally undergoes tetravalent reduction to water. In the ubiquinone cycle, ubiquinone (Q) reduces cytochrome b through several steps, which also result in oxidation of NADH dehydrogenase. This process is also coupled to the transfer of an electron from ubi- quinol (QH2), to cytochrome c1 through iron-sulfur proteins (Fe-S). Superoxide radicals (.O;), and hydrogen peroxide (H~02) can be formed when univalent reduction of oxygen takes place, at the site of interaction of ubiquinone with Fe- S centers. Amytal blocks mitochondrial respiration by inhibiting NADH dehydrogenase. KCN inhibits cytochrome oxidase. Fe-S, iron-sulfur center; QH', ubisemiquinone; cyt, cytochrome. (Figure was adapted from Ref. 11.)

Adytal

conditions previously documented to minimize artifactual formation of oxygen-centered radicals (24). The tissue particles were then poured along with liquid nitrogen through a funnel into precision 3- mm (inner diameter) quartz EPR tubes. Care was taken to maintain the tissue, the funnel, and the EPR tube under liquid nitrogen at all times. EPR spectra were recorded at a temperature of 77 K with a Bruker ESP-300 spectrometer operating at X-band (9.3 GHz) using a TE102 cavity (1). Measurements were performed with non-saturat- ing (ie. 1.0 milliwatt) microwave power. Quantitation of radical concentration was performed from the ratio of the double integral of the observed signal to that of a known concentration of aqueous potassium peroxylamine disulfonate radical s t a n d a r d obtained under identical conditions. The spectra were resolved into individual component signals by computer simulation of the various components, using an EPR simulation program for anisotropic systems (24,31).

NMR Experiments "P NMR spectra were obtained with a Bruker WH 180 spectrom-

eter at 4.23 tesla. At this field strength, phosphorus resonates at 72.89 MHz. The diameter of the probe was 25 mm. The instrument was interfaced to a Nicolet 1280 computer and operated in the pulsed Fourier transform mode. Proton-decoupled spectra were collected over serial 5-min acquisition periods from transients following 45' pulses delivered at 2-8 intervals, conditions previously documented to result in minimal spectral saturation (32). The data were accumulated with a 2,000 table at a 3,000-Hz spectral width. Phosphorus metabolite content was measured by integration of the areas under the individual peaks. Data are expressed as percent of pre-ischemic base- line content. Intracellular pH was determined from the chemical shift (6,) of the inorganic phosphate peak by Equation 1.

To minimize tissue inhomogeneity effects, chemical shifts were measured relative to the phosphocreatine peak, whose position is relatively pH-independent over the range of pH values encountered in this study (pK. = 4.6). The constants wed in this equation are: pK = 6.90; 6~ = 3.290 ppm; & = 5.805 ppm (33).

Biochemical Measurements Tissue Homogenization-Following removal of both atria and the

right ventricle, the left ventricle was homogenized in nine volumes of 1.15% KC1 solution (34). To prevent auto-oxidation of the samples, homogenization was carried out at 4 "C in nitrogen-equilibrated solutions, in the presence of 10 p~ deferoxamine, 0.04% butylated hydroxytoluene, and 2% ethanol (34, 35). The homogenate was initially centrifuged at 1,000 x g for 10 min at 4 "C to remove nuclei

C y i c

1

and tissue debris. The supernatant was centrifuged at 8,000 X g for 30 min at 4 "C. The pellet was resuspended with homogenization buffer, while the supernatant was centrifuged at 30,000 X g for 30 min at 4 "C. The new pellet was resuspended with the homogenization buffer, and the supernatant stored at -70 "C. This supernatant fraction was subsequently thawed and further centrifuged at 100,000 X g for 60 min at 4 "C, and the sediment was resuspended in 2 ml of homogenization buffer. Previous studies in intact hearts (25, 36), as well as in isolated myocytes (37), showed that mitochondria sediment with the 8,000 X g fraction and that the 30,000 X g pellet contains mostly lysosomes, while the microsomal fraction is recovered in the 100,000 X g sediment.

Tissue MDA Assay-Malonyl dialdehyde (MDA)' content was measured on the subcellular fractions by a modified thiobarbituric acid (TBA) method (25, 38, 39). Briefly, 0.6 ml of sodium dodecyl sulfate (8.1%) were added to 0.4 ml of sample. Deferoxamine (4 p ~ ) , butylated hydroxytoluene (0.16%) and ethanol (8%) were added again to the sample to prevent artifactual production of MDA from break- down of lipid peroxides during the assay (34, 35). The sample was then treated with 3 ml of 20% acetic acid and 3 ml of 0.8% TBA and incubated for 90 min at 95 "C in an oil bath. After cooling, the samples were extracted with 2 ml of butanol/pyridine (15/1), and the absorbance read at 532 nm. Calibration curves were prepared daily, using a standard of malonyl dialdehyde tetraethylacetal (which hy- drolyzes to yield MDA). The assay was linear across the 0.2-2.0 nmol/ sample range. In our conditions, recovery of known amounts of MDA standards added to cardiac homogenate8 was >go%, over a broad range of concentrations (0.2-2.0 nmol/sample) (25).

Protein Assay-Protein content of the various fractions was measured by the method of Lowry et al. (40), using bovine serum albumin as a standard.

Oxygen Radical Scavenging-The ability of 5 mM Amytal to act as a scavenger of superoxide radicals or of hydroxyl radicals was evaluated in vitro.

Superoxide radicals were generated in the reaction of xanthine (0.2 mM) with xanthine oxidase (29 milliunits/ml) in 100 mM phosphate buffer, pH 7.2 (41). Superoxide dismutase-inhibitable superoxide radical production was monitored following the rate of reduction of cytochrome c (1.2 mM) at 550 nm, at 37 'C, in a double-beam spec- trophotometer (UVIKON 810, from Kontron Instruments, Zurich, Switzerland) (41).

Hydroxyl radicals were generated using an iron redox system consisting of 1 mM H20* and 20 p~ Fe3+-NTA (1:2) in Dulbecco's phosphate-buffered saline buffer, as described previously (42). Hy-

The abbreviations used are: MDA, malonyl dialdehyde; TBA, thiobarbituric acid; DMPO, 5,5'-dimethyl-l-pyrroline-N-oxide; NTA, nitrilotriacetic acid.

Mitochondria and Reperfuswn Injury 18535

droxyl radical formation was measured by EPR spectroscopy of OH ' adducts formed in the presence of a 50 mM concentration of the spin- trap 5,5'-dimethyl-l-pyrroline-N-oxide (DMPO). EPR spectra were recorded at room temperature using an IBM-Bruker ER 300 spectrometer operating at X-band with a TM 110 cavity and TM flat cell. The spectrometer settings were: modulation frequency, lOq kHz; modulation amplitude, 0.5 G; scan time, 1.0 min; microwave power, 20 milliwatts, microwave frequency, 9.772 GHz. The microwave frequency and magnetic field were precisely measured, respectively, with an EIP 575 Source Locking Microwave Counter and a Bruker ER 035M NMR Gaussmeter.

Glutathione Metabolism-Specific experiments were performed to investigate whether cyanide could interfere with GSH, in vitro or in vivo. In vitro, the reactivity of free -SH groups of GSH was evaluated by incubating varying amounts of reduced glutathione with Ellman's reagent (5,5'-dithiobis-(2-nitrobenzoic) acid, 20 WM in 0.1 M phosphate buffer, pH 7.0, at 25 'C (43). The reaction was monitored following the changes in absorbance at 412 nm, either in the presence or in the absence of 5 mM KCN.

The effects of cyanide on the enzyme of the glutathione cycle were investigated as follows.

Glutathione peroxidase activity was assayed, either in the presence or in the absence of 5 mM KCN, by the method of Lawrence and Burk (44), from the rate of oxidation of NADPH at 22 "C. Varying amounts of glutathione peroxidase (from bovine erythrocytes, specific activity 580 units/mg of protein) were incubated in the presence of 1 mM EDTA, 1 mM NaNa, 0.2 mM NADPH, 1 mM GSH, in 50 mM potassium phosphate buffer, pH 7.0. The reaction mixture contained glutathione reductase (0.8 IU), and was started by adding 2.5 nmol

Glutathione reductase activity was assayed, either in the presence or in the absence of 5 mM KCN, following the oxidation of NADPH after addition of 2.5 nmol of H202 to varying amounta of glutathione reductase (from bakers' yeast, specific activity 190 units/mg of protein) (44). The reaction mixture was identical to that employed for the glutathione peroxidase assay, with the exception that in this case glutathione peroxidase (0.8 UI) was substituted for glutathione reductase.

Finally, the effects of cyanide on tissue glutathione concentration were directly evaluated in three rabbit hearts perfused for 30 min with 5 mM KCN. At the end of perfusion, the hearts were rapidly freeze-clamped between Wallenberg tongs kept under liquid nitrogen. The frozen muscle was then pulverized in a ceramic mortar. Tissue glutathione concentration were then measured by the method of Tietze (45), as previously described (46), and expressed as nmol/mg of protein.

of HzOZ.

Oxygen Consumption Oxygen content was measured by an automated gas analyzer on

samples taken simultaneously from the perfusion buffer going into the heart (via a side arm located just above the tip of the aortic cannula), and from the coronary sinus effluent (collected from a cannula positioned in the right ventricle). Perfusate samples were collected in gas-tight glass syringes, and kept on ice until assayed. Oxygen consumption was calculated from the arterial-venous differ- ence in oxygen content of matched samples and corrected for coronary flow and heart weight (47).

Ventricular Function Experiments Since the effects of Amytal are known to be reversible (48,49), in

an additional series of experiments we tested the hypothesis that transient inhibition of mitochondrial respiration during early reflow by Amytal would also result in improved functional recovery of reperfused hearts after wash out of the drug. Two groups of eight hearts each were studied. One group was treated with Amytal during ischemia and for the first 10 min of reflow, as described above. The infusion was then discontinued, and the hearts were switched to normal perfusate for the remaining 35 min of a 45-min reperfusion period. The intraventricular balloon was left deflated during the initial 15 min of reflow (corresponding to the period of drug infusion and the first 5 min of wash-out), to minimize energy requirements. Control hearts were subjected to a similar protocol. In both groups, functional measurements were obtained at 15, 30, and 45 min of reflow, with the balloon containing the same amount of fluid that had been removed just prior to ischemia. Two Amytal-treated hearts, which fibrillated during reflow, were excluded from the study.

Chemicals Sodium amobarbital (Amytal") was purchased from Lilly. Xan-

thine oxidase (from bovine milk, specific activity 1 unit/mg of protein) was obtained from Boehringer Mannheim GmbH, Mannheim, Ger- many. DMPO was from Aldrich and was purified by double distilla- tion before use. All other reagents were from Sigma.

Statistical Analysis

Data are expressed as mean f S.E. Differences in free radical concentrations and MDA levels among the three groups were tested by one-way ANOVA, followed by Bonferroni-corrected t tests for unpaired data. Differences in the recovery of the various parameters among the groups were tested using a repeated measures two-way analysis of variance (ANOVA). Individual comparisons were performed by t test (at selected time points) only when the overall ANOVA revealed statistical significance ( p < 0.05).

RESULTS

Oxygen Radical Production-Control hearts subjected to ischemia and reperfusion showed the typical EPR signals previously described in this experimental model (1, 4, 24). The overall magnitude of the signal was markedly lower in those heart treated with Amytal (Fig. 2). Cyanide-treated hearts were characterized by a reduction of the magnitude of the signals as compared to the control hearts, although to a lesser extent than Amytal-treated hearts, and by the appear- ance of a broad shoulder on the left of the main signal.

EPR spectra obtained from reperfused hearts include the summation of several radical signals of different concentrations, which can be resolved into individual components (1, 24, 31). By computer processing, four different signals were distinguished (Fig. 3). Signal A was an isotropic signal typical of ubisemiquinone, a carbon-centered radical in the mitochondrial respiratory chain (50); Signal B was an anisotropic signal, typical of an oxygen-centered alkyl-peroxy radical (51, 52); Signal C was a triplet, indicative of a nitrogen-centered radical (1); Signal D was a small signal with two "g" values, attributable to an iron-sulfur protein center of the mitochondrial respiratory chain (53). When the relative contribution

2 Control

d Cyanide

I I 1 I

3100 3200 3300 3400 3500

MAGNETIC FIELD (GAUSS)

three hearts subjected to 30 min of global ischemia at 37 O C FIG. 2. Typical electron paramagnetic resonance spectra of

and 30 s of reflow, under control conditions, or during inhibition of mitochondrial respiration by either 5 mM Amytal or 5 mM KCN (see "Results").


L 1 I I 1 I I I I 3100 3200 3300 3400 3500

Magnetic Field (Gauss) FIG. 3. Resolution of a typical EPR spectrum of a control

heart subjected to ischemia and reperfusion (bottom tracing) into individual components attributable to specific radical species (see "Results"). Signal A, g = 2.004 carbon-centered (ubisemiquinone) radical; signal B, gl= 2.033, g, = 2.005: oxygen-centered (alkyl peroxy) radical; signal C, g = 2.000, hyperfine splitting aN = 24 G nitrogen-centered (nitroxyl) radical; signal D, g, = 2.027, gz = 1.936 iron-sulfur center.

TABLE 1 Effects of mitochondrinl inhibition on free radical concentrations in

reperfused hearts Signal A, carbon-centered (ubisemiquinone) radical; Signal B, oxy-

gen-centered (alkyl peroxy) radical; Signal C, nitrogen-centered (nitroxyl) radical; Signal D, iron-sulfur center. Concentrations are mi- cromolar.

Sienna1 Control Amvtal KCN

A 4.89 f 0.76 2.94 f 0.60" 3.74 f 0.40 B 3.83 f 0.54 1.98 f 0.26* 2.58 f 0.96 C 1.42 f 0.19 1.18 f 0.29 0.96 2 0.29 D 1.78 f 0.12 2.44 f 0.54 3.96 f 0.19'

" p < 0.05 versus control. b p < 0.01 versus control. ' p < 0.001 versus control; p < 0.05 versus Amytal.

of each radical species to the overall EPR spectrum was assessed for each heart, we observed that tissue concentrations of most of these radicals were affected by inhibitors of mitochondrial respiration. Amytal-treated hearts showed significantly lower concentrations of ubisemiquinone radical as compared to controls, consistent with inhibition of mitochondrial respiration at a proximal site (Table I). Tissue levels of oxygen-centered radicals were also significantly reduced in reperfused hearts treated with Amytal. Inhibiting mitochondrial respiration at a distal site by cyanide induced lesser reduction in tissue concentrations of the carbon-centered and oxygen-centered radical signals, which did not achieve statistical significance. On the other hand, administration of cyanide, but not Amytal, was associated with a significant increase in the signal from the iron-sulfur center, consistent with the accumulation of this species in its reduced form as a consequence of the distal block of electron flow induced by KCN (Table I).

Metabolic Effects of Ischemia-In control hearts, 30 min of global ischemia at 37 "C induced a rapid depletion of phosphocreatine and a more gradual decrease of ATP content, as assessed by NMR spectroscopy (Table 11). These effects were accompanied by progressive accumulation of inorganic phosphate and by development of intracellular acidosis (Table 11).

TABLE I1 Effects of mitochondrial inhibition on phosphorous metabolite content

at the end of ischemia Data obtained by "P NMR spectroscopy at base-line level (100%)

and at the end of 30 min of 37 'C ischemia. Differences among groups were not statistically significant.

Control Amytal KCN

Phosphocreatine (% of 3.6 f 1.4 3.6 rt 0.8 6.3 f 1.5

ATP (% of base-line) 39.5 * 4.1 50.1 f 9.0 46.2 f 4.7 Inorganic phosphate 1,115 f 131 1,244 f 146 961 f 114

Intracellular pH

base-line)

(% of base-line)

Base-line level 7.11 f 0.03 7.08 f 0.04 7.10 f 0.03 End of ischemia 6.30 f 0.10 6.46 f 0.10 6.21 f 0.16

Severity of ischemia was similar in the three groups. Membrane Lkid Peroxidation-Accumulation of MDA, an

index of oxygen radical attack of membrane lipids (25, 34, 39), was significantly reduced in the mitochondrial fraction of those hearts that received Amytal, as compared to either untreated hearts or hearts treated with cyanide (Fig. 4, top panel). MDA levels were also lower in the lysosomal fraction of hearts treated with Amytal (p < 0.05 versus cyanide-treated hearts; p < 0.05 versus control hearts, by corrected t test) (Fig. 4, middle panel). Malonyl dialdehyde levels in the microsomal fraction were higher than controls in Amytal-treated hearts (p < 0.05) and in the group that received KCN (p = NS) (Fig. 4, bottom panel).

Oxygen Consumption-The three groups of hearts had similar values of O2 consumption during the pre-ischemic base- line period. In control hearts, O2 consumption was transiently elevated in the first minute of reflow and then remained slightly depressed during the remainder of postischemic reperfusion (Table 111). In Amytal-treated hearts, oxygen consumption was markedly depressed during the initial 10 min of reperfusion with the drug. Inhibition of 0 2 consumption by Amytal was similar to that reported in previous studies (48, 49,54), and it has been calculated that under these conditions mitochondrial respiration is inhibited by >95% (48, 54). 0 2

consumption recovered to values similar to those of untreated hearts following discontinuation of the infusion of Amytal (Table 111), consistent with the effects of Amytal being fully reversible (48,49).

Effects of Amytal on Oxygen Radicals-Experiments were performed to determine whether at the concentrations employed Amytal could significantly scavenge oxygen radicals. Superoxide dismutase-inhibitable production of superoxide radicals by the xanthine/xanthine oxidase system averaged 21.8 f 3.9 nmol/min under control conditions, and it was 18.9 f 1.4 nmol/min in presence of 5 mM Amytal (p = NS). Possible scavenging effects on hydroxyl free radicals were then investigated by EPR evaluation of DMPO-OH adducts in the H202/FeS+-NTA reaction. In the absence of HzOz, only trace signals were observed. However, upon addition of hydrogen peroxide, a strong 1:2:2:1 quartet signal with aH = aN = 14.9 G (indicative of DMPO-OH) was seen (Fig. 5A). When 5 mM Amytal was present, an identical signal was observed. Intensity of the DMPO-OH adduct signal was not decreased by Amytal (Fig. 5B). As described previously (42), in the presence of 1% ethanol this system gives rise to the hydroxy- ethyl adduct, confirming that the observed DMPO-OH adduct arises from the trapping of 'OH. Therefore, at the concentration employed to perfuse the hearts, Amytal does not significantly scavenge superoxide radicals or hydroxyl radicals.

Effects of Cyanide on Glutathione-Cyanide may react with

Mitochondria and Reperfusion Injury 18537

free -SH groups. Therefore, we also tested whether lack of protection from KCN might have been due to a concomitant interference with cardiac GSH (a major intracellular system to inactivate oxidants), since this effect would have acted to cancel out benefits possibly deriving from mitochondrial inhibition by cyanide. At the dose employed in the present

' 3 Mitochondrial Fraction

n.x 1 1 .... I

0.6 - * # T

0.4

0.2 :

O J Controls Amytal KCN

Lvsosomal Fraction

1

0.4

0.2 k Controls Amytal KCN

'7 Microsomal Fraction

Controls Amytal KCN FIG. 4. Effects of inhibiting mitochondrial respiration on

the concentration of lipid peroxidation products in cardiac membranee. Hearts were subjected to 30 min of global ischemia at 37 "C, and 10 min of reflow, under control conditions (open bars), or during treatment with Amytal (cross-hatched bars) or KCN (closed bars). Lipid peroxidation was measured from MDA content, expressed as nanomoles of TBA-reactive product, and normalized for protein content. *, p < 0.05 uers'sus controls (by Bonferroni-corrected t test); # , p < 0.05 uersus KCN (by Bonferroni-corrected t test).

study, cyanide had no effects on glutathione metabolism. In fact, the number of 5,5'-dithiobis-(Z-nitrobenzoic) acid-reactive SH- groups of GSH was not affected by the presence of a large excess of KCN (5 mM), over a wide range of GSH concentrations (Fig. 6, upper panel). Similarly, neither glutathione reductase nor glutathione peroxidase showed any decrease in activity when incubated in the presence of 5 mM KCN (Fig. 6, middle and lower panels). Finally, tissue GSH concentrations averaged 8.5 f 1.2 nmol/mg of protein in three hearts that were perfused for 30 min with 5 mM KCN, similar to the value of 9.7 f 0.9 nmol/mg of protein, found in three hearts perfused with control buffer.

Effects of Amytal on Recovery of Ventricular Function- Thirty min of ischemia produced a marked impairment in recovery of postischemic left ventricular function in control hearts. In this group, left ventricular developed pressure at the end of 45 min of reperfusion averaged 48 * 5% of pre- ischemic base-line level. In contrast, those hearts that received Amytal during the initial 10 min of reflow showed a significantly greater recovery of contractility. In this group, left ventricular developed pressure averaged 66 f 8% of base- line levels 15 min after reflow (uersus 43 & 8% in controls; p = 0.09), and 67 * 7% at the end of the 45-min reperfusion period ( p < 0.05 versus controls; Fig. 7, left panel). Similarly, impairment of diastolic function during reperfusion was significantly more pronounced in controls compared to Amytal- treated hearts ( p c 0.05 by ANOVA and p < 0.05 at 15 min) (Fig. 7, right panel). There was also a tendency toward higher coronary flows during reperfusion in the group treated with Amytal (75 f 6% of base-line levels at 45 min of reperfusion uersus 68 f 6% in controls) ( p = NS).

DISCUSSION

In the present study, administration of the mitochondrial respiration inhibitor, Amytal, to intact hearts reperfused after a period of ischemia was associated with reduced formation of oxygen radicals at reflow, as compared to hearts reperfused under control conditions. This phenomenon was also accompanied by reduced membrane lipid peroxidation in the mitochondria as well as in the lysosomes isolated from Amytal- treated hearts subjected to ischemia and reflow. These effects were not seen when mitochondrial respiration was blocked by cyanide, which, unlike Amytal, acts at the distal end of the respiratory chain. Finally, transient inhibition of mitochondria by Amytal was associated with greater functional recovery after ischemia and reperfusion, as compared to untreated hearts. Taken together, these results demonstrate that mitochondrial generation of oxygen radicals can occur in intact hearts and indicate that resumption of mitochondrial respiration might be an important contributor to tissue damage following postischemic reperfusion.

It has long been known that superoxide radicals and hydrogen peroxide can be generated during mitochondrial respiration, and it has been hypothesized that this phenomenon may contribute to tissue alterations associated with aging (55,56), hyperbaric oxygen toxicity (lo), and reperfusion injury (5, 6,

TABLE I11 Effects of mitochondrial inhibition on cardiac oxygen consumption

Data are expressed as microliters of O2 consumed per minute per gram, wet weight (mean f S.D.).

Base-line level

Controls Amytal

0.070 f 0.011 0.079 f 0.011 0.059 f 0.010 0.054 f 0.001 0.064 f 0.013 0.068 f 0.011 0.020 & 0.002

KCN 0.015 f 0.004 0.016 f 0.002 0.056 f 0.006

0.072 f 0.010 0.019 f 0.004 0.006 f 0.001 0.007 f 0.001 0.028 f 0.017

Reperfusion

1 min 5 min 10 min 20 min


Magnetic Field (Gauss)

FIG. 5. EPR spectra of the iron redox hydroxyl radical generating system consisting of 1 mM HaO$ and 20 NM FeS+- NTA, in the presence of 50 m~ DMPO. A series of 1-min EPR acquisitions were recorded. A, EPR spectrum observed after 4 min of reaction under control conditions; B EPR spectrum observed after 4 min in the presence of 5 mM Amytal.

5 mM KCN

0 2 4 6 8 1 0 1 2 1 4

[GSHI nhloledml 0.15 7

0 0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 Glutathione Peroxidase (Unidml)

0.08 1

0.00 4 , 1

0.0 0.5 1.0 IS 2.0 2.5 3.0 3.5 4.0 Glutathione Reductase (Unidml)

FIG. 6. Upper p a n e l , effects of 5 mM KCN on -SH group availability of reduced glutathione, as determined by reaction with Ellman's reagent (see "Materials and Methods"). Middlepanel, effects of 5 mM KCN on the activity of glutathione peroxidase. Lower panel , effects of 5 mM KCN on the activity of glutathione reductase.

21). In this respect, we have recently observed that, in hearts reperfused after a period of ischemia, the time course of oxygen radical generation closely mirrors the changes in intensity of the signal originating from the 1-electron reduced ubiquinone of the mitochondrial respiratory chain (4). This observation would suggest that the burst of mitochondrial respiratory activity that occurs when oxygen is reintroduced upon postischemic reflow might be responsible for the simul- taneous burst of oxygen radical formation on reperfusion. However, although oxygen radical formation has been demonstrated repeatedly in in vitro studies employing isolated mitochondria (10-17, 21, 57, 58), occurrence of this phenomenon in intact organs has not been documented previously. Our data for the first time demonstrate a direct link between mitochondrial respiration and generation of oxygen-centered free radicals in intact hearts.

Under physiologic conditions, small quantities of oxygen radicals that may form during mitochondrial respiration can be detoxified by endogenous scavenging mechanisms of myocardial cells (5, 10). However, postischemic reperfusion may alter this balance. On the one hand, a decrease in the cellular levels of superoxide dismutase and other endogenous scavengers has been observed in postischemic hearts (18-20,57). On the other hand, the ischemic episode may cause specific mitochondrial changes, which would favor oxygen radical production when oxidative phosphorylation resumes upon reflow. In this respect, in vitro studies have documented that mitochondrial formation of oxygen radicals mostly occurs during State 4 respiration, i.e. when mitochondria are provided with oxygen and substrates from the Krebs cycle, but the electron flow is limited by lack of ADP to phosphorylate. Under this condition, oxygen radical production in vitro is further enhanced by interventions that shift the components of the respiratory chain toward a more reduced state (11,13,16,17). A similar situation of ADP depletion and altered redox state is likely to be encountered at the time of reflow, when oxygen is reintroduced to mitochondria which have accumulated large quantities of reducing equivalents during ischemia. These effects of ischemia on the redox state of mitochondria have been demonstrated in intact rat hearts by means of direct EPR spectroscopy by Baker and Kalyanaraman (531, who observed a progressive reduction of the iron-sulfur centers of the electron transport chain with the onset and development of ischemia. A similar phenomenon has been also observed in the in vivo rabbit heart by Grill et al. (59). Furthermore, when oxygen supply is restored during reperfusion, electron transport through the respiratory chain may be further impaired due to the depletion of ADP during ischemia. This may increase the transfer of electrons directly to molecular oxygen to generate superoxide radicals.

Amytal is known to act by inhibiting the oxidation of various substrates by mitochondrial NADH dehydrogenase (28). This enzyme is located at Site I of the respiratory chain, proximal to ubiquinone (Fig. 1). Thus, the effects observed in the present study when mitochondrial respiration was blocked by Amytal would extend to intact hearts the notion derived from studies performed on isolated mitochondria that oxygen radicals can be generated in the mitochondrial chain at the sites of interaction between Fe-S clusters and ubiquinone, and that NADH dehydrogenase is a major source (11,15-17) (Fig. 1). In addition to the classical pathway of oxygen radical generation depicted in Fig. 1, it is also possible that in postischemic hearts NADH dehydrogenase activity may produce superoxide radical by an alternative mechanism. McCord and associates (60) have recently proposed that, in mitochondria isolated after a period of cardiac anoxia and reoxygenation,


FIG. 7. Effects of transient inhibition of mitochondrial respiration - by Amytal on recovery of isovol- f 1 umic left ventricular developed E pressure (left panel) and of end-di- 1. astolic pressure (right panel). The p overall time course of both parameters E treated hearts ( p < 0.05 by ANOVA) was significantly different in Amytal- Jo :

d:

first 10 min of reflow). mM Amytal (during ischemia and for the Heperfusion ’l‘imc (minl Reperfusion Time (min) received either 5 mM KC1 (controls) or 5 0 I5 Jn 45 0 15 30 45

. . I 1 2 o;. . . . , . . . , . . . . , chemia at 37 “C, and 45 min of reflow, - 0. sons). Hearts subjected to 30 min of is- 2 9 (see “Results” for individual compari- = y 20 : 1

IOU 7 + Controls a loo f X

t Amytal E

L 60 - .Y 40

20 - Q < 0.05 p < 0.05

NADH dehydrogenase becomes highly reduced and may transfer electrons directly to oxygen (instead of ubiquinone), generating * 0;.

The results of the present study also indicate that inhibiting mitochondrial respiration by KCN did not significantly reduce myocardial production of oxygen radicals. This finding is not entirely surprising. In uitro, cyanide has been shown to block the formation of superoxide radicals and Hz02 at the ubiquinone-cytochrome b site (16). However, it has also been shown that KCN does not affect superoxide radical generation at the NADH dehydrogenase site (16,61), and it may in fact enhance it, as cyanide induces full reduction of the respiratory chain components (11,16,61). In this respect, it is interesting that in our study the signal originating from reduced Fe-S centers was markedly elevated in the group of hearts treated with cyanide, consistent with greater reduction of respiratory chain components induced by this agent.

In the present study, inhibition of mitochondrial oxygen radical generation also exerted potentially relevant biological effects. Amytal treatment was accompanied by reduced accumulation of lipid peroxidation products at the level of mitochondrial and lysosomal membranes. Lipid peroxidation secondary to oxygen radical attack to double bonds of membrane fatty acids has been proposed as an important mechanism of cell damage. It may alter intrinsic membrane properties, due to physicochemical changes of oxidized lipids, or secondary to cross-linking and polymerization of membrane components effected by MDA (62, 63). In this respect, we have recently shown in this same experimental model that postischemic reperfusion is associated with a significant increase in the levels of malonyl dialdehyde and of conjugated dienes (another marker of lipid peroxidation) in cardiac mitochondria and lysosomes, which was specifically prevented by oxygen radical scavengers (25). Increased tissue levels of malonyl dialdehyde and/or conjugated dienes in reperfused hearts have been found by other investigators (29, 30, 64). Thus, the parallel changes observed between oxygen radical concentration and lipid peroxidation in the present study strongly suggest that oxygen radicals formed in the process of mitochondrial respiration may contribute to some of the alterations that have been associated with reperfusion injury. This point is directly supported by the second series of experiments, showing improved functional recovery with Amytal treatment during early reflow. Additional support to the hypothesis that resumption of mitochondrial respiration can exert deleterious effects on the heart comes from earlier observations by Hearse and Humphrey (see Ref. 65). In a different experimental model, they observed that cardiac release of creatine kinase (a marker of loss of integrity of cell membranes) in hearts reoxygenated after a period of anoxic perfusion (i.e. “oxygen paradox”) was markedly reduced by Amytal.

In the present study, oxygen consumption fell to about 20%

of base-line level in hearts reperfused with Amytal. Previous studies with this drug in intact hearts have obtained similar results (48, 49)) and it has been estimated that this corresponds to >95% inhibition of mitochondrial respiration (48, 54). In spite of this fact, oxygen radical production was reduced only about 50% by Amytal treatment, even though two other major sources of oxygen radicals are minimally present or absent in our model (i.e., xanthine oxidase and neutrophils). This finding indicates that mitochondrial respiration is not the only mechanism of oxygen radical formation in our model. Cyclooxygenase and microsomal oxidases are other potential sources of oxygen radicals in the heart (66, 67). In addition, recent experiments by Vandeplassche et al. (68,69) have documented histochemically that a specific form of NADH oxidase associated with mitochondria (but unrelated to oxidative phosphorylation) may become activated in hearts subjected to ischemia and reperfusion. Interestingly, the activity of this enzyme is known to be unaffected by mitochondrial inhibitors, and it is actually expected to be stimulated when the mitochondrial respiratory chain is blocked (70). Therefore, this metabolic pathway might have remained operative in spite of Amytal administration, and in fact it might be speculated that the drug, by preventing NADH and oxygen from being utilized in the oxidative phosphorylation, might have actually increased the amount of oxygen radicals generated by way of NADH oxidase, as well as by microsomes (67, 71) and/or other sources. A similar mechanism might also explain the finding in the present study of lipid peroxidation being more pronounced in the microsomal fraction of Amytal-treated hearts. Again, it is possible that microsomal oxidase activity may be more pronounced when 0, utilization is diverted from the mitochondrial respiration process.

The present study has certain limitations. One obvious limitation is represented by the fact that our experimental conditions do not allow to perform a detailed characterization of the relationship between mitochondrial respiration and oxygen radical production, which can be optimally performed on isolated mitochondria, and not in intact beating hearts. Another possibility is that the beneficial effects of Amytal that we observed might have been unrelated to inhibition of mitochondrial respiration. However, Amytal has been previously used to specifically block mitochondrial respiration in intact hearts (48, 49, 54). In fact, detailed analysis of the relationship between mitochondrial respiratory rates and Amytal concentrations indicates that in the perfused heart this drug affects only a single component (i.e. NADH dehydrogenase) (48). In addition, our in uitro experiments documented that Amytal was devoid of direct scavenging effects on superoxide radicals. Finally, the 31P NMR data allowed us to confirm that Amytal had no effects on the severity of ischemia. Another factor to be considered when interpreting


the results of the present study is the experimental model. It is well known that production of oxygen radicals increases with increasing the oxygen tension to which isolated mitochondria are exposed in uitro (13, 67, 71). Therefore, it might be hypothesized that the high oxygen tension of crystalloid perfusate might have enhanced this phenomenon in our ani- mal model, whereas in vivo other mechanisms might account for a larger fraction of oxygen radical generation. Against this speculation, however, is the fact that, in spite of higher oxygen tension, 02 carrying capacity of crystalloid perfusates is actually lower than that of blood. Furthermore, in intact hearts intracellular O2 tension ut the mitochondrial leuel is markedly lower than in the vascular lumen (72), and presumably similar both in uiuo, and in in uitro perfused hearts.

Another potential concern relates to possible artifacts in the measurement of free radical concentrations or of lipid peroxidation products. In the present study, free radicals were measured by direct EPR spectroscopy of frozen tissue. It has been reported that oxygen-centered radicals may be artifac- tually generated during preparation of the sample for this type of assay (24, 73, 74). However, we have previously characterized this process and demonstrated that this problem can be minimized through appropriate sample handling (24). Mechanical processing of tissue samples was shown to result in the formation of alkyl radicals (R') through cleavage of covalent bonds. In the presence of oxygen, these R' radicals react to form ROO'. Thus, artifactual formation of oxygen- centered signals can be minimized by reducing to a minimum the mechanical fracturing of the tissue and by rigorously maintaining the sample under anaerobic conditions throughout the procedure. By this approach we observed that the results obtained in intact hearts were identical to those achieved by EPR spectroscopy of cardiac specimens directly analyzed without tissue grinding, i.e. as obtained by needle biopsy or by high speed drill biopsy (24, 59). Furthermore, all features of reperfusion-induced oxygen radical generation, including time course, oxygen dependence, response to scavengers, role of iron, were consistently observed, in a similar fashion, when oxygen radicals were measured either by direct EPR spectroscopy of frozen tissue or by indirect spin-trap methodology (which does not require tissue processing) (1, 2, 4, 24, 75, 76). Thus, with proper care and controls, valid measurements of free radicals can be performed by direct EPR spectroscopy. In fact, this technique may offer some advantages over the spin-trap methodology when investigat- ing the specific issues of the present study. On the one hand, perfusion of the heart with spin-traps does not allow rapid and complete cell permeation with the agent, and therefore oxygen radicals produced immediately upon reflow by an intracellular mechanism (such as mitochondria) might go undectected. On the other hand, only direct EPR spectroscopy of frozen tissue can detect changes in the redox state of components of the mitochondrial respiratory chain that are not oxygen-centered, namely ubiquinone and Fe-S centers.

Chemical measurements of tissue concentrations of MDA also have certain limitations. Artifactual increases in MDA content may be due to chemical interference from other TBA- reactive substances present in the sample or to subsequent auto-oxidation causing degradation of lipid peroxides during sample handling. These problems can be reduced by proper assay procedures (34, 35, 64, 77), and in fact MDA is widely utilized as a marker of lipid peroxidation both in uiuo and in perfused hearts (25, 38, 39, 78, 79). On the other hand, measurement of MDA content may underestimate the actual degree of lipid peroxidation, since MDA is formed only when lipids containing three or more double bonds are oxidized.

This latter phenomenon might explain why the protective effects of Amytal on MDA accumulation are of a smaller magnitude as compared to the inhibitory effects we observed on oxygen radical production.

In conclusion, reversible inhibition by Amytal of electron transport at Site I during early reflow can reduce oxygen radical production, decrease membrane lipid peroxidation, and at the same time improve functional recovery of rabbit hearts subjected to ischemia and reperfusion. Taken together, these lines of evidence demonstrate that oxidative phosphorylation is a source of oxygen radicals in intact hearts and suggest that resumption of mitochondrial respiration upon postischemic reperfusion might be an important contributor to the pathogenesis of reperfusion injury.

Acknowledgments-We acknowledge the technical assistance of Uriah Lee Shang, Koenraad Vandegaar, and Annalisa Scognamiglio.

REFERENCES 1. Zweier, J. L., Flaherty, J. T., and Weisfeldt, M. L. (1987) Proc. Natl. Acad.

2. Garlick, P. B., Davies, M. J., Hearse, D. J , and Slater, T. F. (1987) Circ.

3. Bolli, R., Patel, B. S., Jeroudi, M. O., Lai, E. K., and McCay, P. B. (1988)

4. Ambrosio, G.. Zweier. J. L., and Flahertv, J. T. (1991) J. Mol. Cell Cardiol.

Sci. U. S. A. 8 4 , 1404-1407

Res. 61,757-760

J. Clin. Inuest. 8 2 , 476-485

23,1359-i374 . .

5. Hess, M. L., and Manson, N. H. (1984) J. Mol. Cell CardwL 16,969-978 6. Becker, L. C., and Ambrosio, G. (1987) Prog. Cardwuasc. Dis. 30.23-41

8. Chambers, D. E., Parks, D. A., Patterson, G., Roy, R., McCord, J., Yoshida, 7. McCord, J. M. (1985) N. Engl. J. Med. 312,159-163

S., Parmley, L. F., and Downey, J. M. (1985) J. MOL Cell. Cardwl. 17,

9. Fantone, J. C., and Ward, P. C. (1982) Am. J. Pathol. 107,394-418 10. Chance, B., Sies, H., and Boveris, A. (1979) Physiol. Reu. 69! 527-605 11. Turrens, J. F., and McCord, J. M. (1990) in ,Clinical Ischemrc Syndromes:

Mechanisms and Consequences of %sue In ury (Zelenock, G. B., D'Alecy, L. G., Fantone, J. C., 111, Shlafer, M., and Stanley, J. C., e&) pp. 203- 212, C. V. Mosby Co., St. Louis, MO

145-152

12. Loschen, G., Floh6, L., and Chance, B. (1971) FEBS Lett. 18,261-264 13. Boveris, A., and Chance, B. (1973) Biochem. J. 134,707-716 14. Lozhen, G., Azzi, A,, Richter, C., and Floh6, L. (1974) FEBS Lett. 4 2 , s

15. Cadenas, E., Boveris, A., Ragan, C. I., and Stoppani, A. 0. M. (1977) Arch.

17. Turrens, J. F., Alexandre, A., and Lehninger, A. L. (1985) Arch. Biochem. 16. Turrens, J. F., and Boveris, A. (1980) Biochem. J. 191,421-427

I Z

Bwchem. Bwphys. 180,248-257

18. Julicher, R. H. M., Tijburp L. B. M., Sterrenberg, L., Bast, A., Koomen, J. Bwphys. 237,408-414

M.. and Noordhoek. J. 1984) M e SCL. 36.1281-1288 19. Fer&,-R.,Ceconi,C.,'Curello, S., &uamieri,C., Caldarera, C. M., Albertini,

A., and Visioli, 0. (1985) J. Mol. Cell. CardwL 17,937-945 20. Arduini, A,, Mezzetti, A., Porreca, E., Lapenna, D., DeJulia, J., Marzio, L.,

Polidoro, G., and Cuccurullo, F. (1988) Biochim. Bmphys. Acta 970,113-

21. Otani H., Tanaka H. Inoue, T., Umemoto, M., Omoto, K., Tanaka, K., 121

Sa&, T., Osako,'T., Masuda, A,, Nonoyama, A,, and Kagawa, T. (1984) Circ. Res. 66,168-175

22. Ambrosio, G., Weisfeldt, M. L., Jacobus, W. E., and Flaherty, J. T. (1987) Circulation 76,282-291

23. Ambrosio, G., Santoro G., Tritto, I., Elia, P. P., Duilio, C., Scognamiglio, A., Basso, A., and Chiariello, M. (1992) Am. J. PhysioL 2 6 2 , H23-H30

24. Zweier J. L., Kuppusamy, P., Williams, R., Rayburn, B. K., Smith, D., Weiifeldt, M. L., and Flaherty, J. T. (1989) J. Bioi. Chem. 264,18890-

25.

26.

27.

29. 28.

30.

31. 32.

33.

35. 34.

36. 37.

38. 39.

40.

Ambrosio, G., Flaherty, J. T., Duilio, C., Tritto, I., Santoro, G., Elia, P. P., 18895

Grum. C. M.. Ragsdale. R. A., Ketai, L. H.. and Shlafer, M. (1986) Bmhem. Condorelli, M., and Chiariello, M. (1991) J. Clin. Znuest. 8 7 , 2056-2066

de Jong, J. W., van der Meer, P., Selma Nleukoop, A., Huizer, T., Stroeve,

Romakhin, A. D., Rebeyka, I., Wilson, G. J., and Mickle, D. A. G. (1987) Slater E. C. (1967) Methods Enzymol. 10,48-57

Biobhys. Res. commun. Ikl,1104-1108,

R. J., and Bos, E. (1990) Circ. Res. 67, 770-773

Romaschin A. D., Wilson, G. J., Thomas, U., Feitler, D. A,, Tumiati, L., J. Mol. Cell. Cardwl. 19, 289-302

and Micile, D. A. G. (1990) Am. J. Physiol. 269, Hll6-Hl23

Flaherty, J. T., Weisfeldt, M. L., Bulkley, B. H., Gardner, T. J., Gott, V. Nettar, D., and Villafranca, J. J. (1985) J. Magn. Reson. 64,61-65

Jacobus, W. E., Pores, I. H., Taylor, G. J., Nunnaly, R. L., Hollis, D. P.,

Slater, T. F. (1984) Methods Eruymol. 106,283-293 Bird, R. P., and Draper, H. H. (1984) Methods EnzynwL 106,299-305 Welma, E., and Peters, T. J. (1976) J. Mol. Cell. CardwL 8,443-463 Weglicki, B. W., Owens, K., Kennet, F. F., Kessner, A,, Harris, L., Wise,

Ohkawa, H., Ohishi, N., and Yaki, K. (1979) An@. Bmhem. 96,351;358 R. M., and Vahouny, G. V. (1980) J. BioL Chem. 266,3605-3609

Liedtke. A. J.. Mahar. C. 8.. Ytrehus, K., and MJOS, 0. D. (1984) Baslc Res.

L., and Jacobus, W. E. (1982) Circulatwn 66,561-571

and Weisfeldt, M. L. (1978) J. Mol. Cell. Cardwl. 10 , 39-46

~~

CaFdibL 79,513-518 '

. . . .

Lawry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951) J. Bwl. Chem. 193,265-275

41. 42.

43. 44.

45. 46.

47.

48.

49.

60.

62. 61.

63. 54.

56.

56. 67.

58.

Mitochondria and Reperfusion Injury 18541

Josephson, R. A., Silverman, H., Lakatta, E. G., Stern, M. D., and Zweier, Mc Cord, J. M., and Fridovich, I. (1968) J. Biol. Chem. 243,5753-5760

Eyer, P., and Podhradsky, D. (1986) A d . Bwchem. 1 6 3 , 5 7 4 Lawrence, R. A., and Burk, R. F. (1976) Bwchem. Biophys. Res. Commun.

Tietze, F. (1969) A d Biockm. 27,502-522 Ferrari, R., Ceconi, C., Curello, S., Guamieri, C., Caldarera, C. M., Albertini,

Renlund, D. G., Lakatta, E. G., Mellits, E. D., and Gerstenblith, G. (1985)

Nishiki, A., Erecinska, M., and Wilson, D. F. (1979) Am. J. Physiol. 237,

Ku riyanov, V. V., Lakomkin, V. L., Korchazhkina, 0. V., Stepanov, V. A., &eemschnelder, A. I. A., and Kapelko, V. I. (1991) Biochim. Biophys. Acta 1058,386-399

Ohnishi, T., and Tnunpower, B. (1980) J. Biol. Chem. 266,3278-3284

Knowles, P. F., Gibson, J. F., Pick, F. M., and Bray, R. C. (1969) Biochem. Copeland, E. S. (1975) J. Mogn. Reson. 20,124-129

Baker, J. E., and Kalyanaraman, B. (1989) FEBS Lett. 244,311-314 Hoerter, J. A., Lauer, C., Vaasort, J., and Guhron, M. (1988) Am. J. Physiol.

Nohl, H., Breuninger, V., and Hegner, D. (1978) Eur. J. Biochem. 90,385-

Nohl, H., and Hegner, D. (1978) Eur. J. Bioehem. 82,563-567 Shlafer, M., Myers, C. L., and Adkins, S. (1987) J. Mol. CeU. Cardiol. 19,

Shlafer, M., Gallagher, K. P., and Adkins, S. (1990) Basic Res. Cardiol. 86,

J. L. (1991) J. BioL C k m . 266,2354-2361

71,952-968

A., and Visioli, 0. (1985) J. Mol Cell Cardiol. 17,937-945

Circ. Res. 67,876-888

C221-C230

J. 11 1,53-58

266, C192-C201

390

1196-1206

3 1 ~ 2 9

60. Turrens. J. F.. Beconi. M.. Barilla. J.. Chavez. U. B.. and McCord. J. M. M. L. (1992) J. Am. CoU. CardioL 20,1604-1611

(1991) FreekadicalRes.’Co&muk. 12-13,681-689 ~ ~ ~ ~~’ 61. Takeshige, K., and Minakami, S. (1979) Bioehem. J. 180, 129-135 62. Hocestem, P., and Jain, S. K. (1981) Fed. Proc. 40.183-192 63. Nielsen, H. (i978) Lip& 16,215-223 64. Koller. P. T.. and Berman. S. R. (1989) Circ. Res. 66.838-846 65. He&, D. J.’ (1977) JrMol.’CeU. Cardwl. 9,605316 ’

66. Freeman B. A and Crapo J. D. (1982) Lab. Inuest. 47,412-426 67. Turrens,’J. F., Freeman, B.’A., and Crapo, J. D. (1983) in Ox Radicals and

Their Scaue er System (Greenvald, R. A., and Cohen d e&) Vol. 2,

68. Vandeplassche, G., Hermans, C., Thon6, F., and Borgers, M. (1989) J. Mol pp. 365-370,%lsevier Science Publishers B.V., Amsterdam”

CeU. Cardml 21,383-392 69. Vandeplassche, G., Hermans, C., Thon6, F., and Borgers, M. (1991) Car-

~~

dioscience 2 47-53 70. Nohl, H. (198;) FEBS Lett. 214,269-273 71. Turrens, J. F., Freeman, B. A., and Crapo, J. D. (1982) Arch. Biochem.

72. Gayeski, T. E. J., and Honig, C. R. (1991) Am. J. Physiol. 260, H522- Biophys. 217,411-421

HF.31 73. Bsir:-J. E., Felix, C. C., Olinger, G. N., and Kalyanaraman, B. (1988)

74. Nakazawa, H., Ichmori, K., Shinozaki, Y., Okino, H., and Hori, S. (1988)

75. Zweier, J. L., Rayburn, B. K., Flaberty, J. T., and Weisfeldt, M. L. (1987)

Proc. Natl. Acad. Sci. U. S. A. 86,2786-2789

Am. J. Physiol. 266, H213-H215

J. Clin. Inuest. 80. 1728-1734 76. Zweier, J. L. (1988) >.-Bid. Chem. 263,1353-1357

78. Paller, M . S., Hoidal, J. R., and Feme., T. F. (1984) J. C l h Inuest. 74, 77. Fantini G. A., and Yoshloka, T. (1992) Am. J. Physiol. 2 6 3 H981-H982

llCC-llCA

Evidence that mitochondrial respiration is a source of potentially toxic oxygen free radicals in...

Documents

Transcript of Evidence that mitochondrial respiration is a source of potentially toxic oxygen free radicals in...