Blocking Free Radical Production via Adenoviral Gene Transfer Decreases Cardiac Ischemia–Reperfusion Injury

ARTICLE

doi:10.1006/mthe.2000.0193, available online at http://www.idealibrary.com on IDEAL

Blocking Free Radical Production via Adenoviral Gene Transfer Decreases Cardiac Ischemia–Reperfusion Injury Henry L. Zhu,* Allan S. Stewart,* Matthew D. Taylor,* C. Vijayasarathy,† Timothy J. Gardner,* and H. Lee Sweeney‡,1 *Department of Surgery and ‡Department of Physiology, School of Medicine, and †Department of Animal Biology, School of Veterinary Medicine, University of Pennsylvania, Philadelphia, Pennsylvania 19104 Received for publication June 2, 2000, and accepted in revised form September 29, 2000

Periods of cardiac ischemia followed by reperfusion can lead to either transient loss of function (stunning) or permanent functional loss stemming from infarction, depending upon the length of the ischemic period. In either case the primary mediator of the injury may by oxygen-derived free radicals generated upon the reestablishment of blood flow. The heart’s primary defense against peroxide, glutathione peroxidase, is depleted during ischemia. Thus, the ischemic myocardium might derive significant protection from increased levels of the enzyme, catalase, which can remove hydrogen peroxide in a redox-independent manner. To test these assertions, we studied the ability of adenoviral gene transfer to increase intracellular antioxidant activity via catalase expression. What we observed was that increasing catalase activity in the heart was sufficient to prevent the stunning associated with 15 min of ischemia followed by reperfusion. Key Words: free radicals; antioxidants; catalase; reperfusion.

INTRODUCTION Periods of transient ischemia followed by reperfusion can lead to cellular damage and functional loss in a variety of tissues. This is particularly a problem for the heart, where there is considerable evidence that additive myocardial injury occurs during reperfusion of ischemic myocardium (1, 2). The longer the ischemic period, the greater and less reversible the damage that occurs upon reperfusion. For example, short periods (minutes) of ischemia can result in transient, reversible depression of contractile function (known as stunning) upon reperfusion. More protracted periods of ischemia, however, lead to extensive myocyte death and regional infarction. Reactive oxygen species generated during reperfusion may play an important role in the initiation of reperfusion injury (3, 4). Detoxification of H2O2 in the myocardium is mainly accomplished by the glutathione peroxidase enzyme system. However, glutathione (GSH) is depleted during myocardial ischemia (5, 6) and reperfusion (7, 8). Therefore, the capacity for H2O2 degradation is greatly diminished in the heart immediately following a period of ischemia. In many other tissues, catalase is the major

1To whom correspondence and reprint requests should be addressed at Department of Physiology, University of Pennsylvania School of Medicine, A-700 Richards Building, 3700 Hamilton Walk, Philadelphia, PA 19104-6085. Fax: 215-898-0475. E-mail: [email protected].

470

antioxidant enzyme that converts hydrogen peroxide (H2O2) to water and oxygen and prevents the formation of the highly reactive hydroxyl radical (•OH); however, catalase levels are low in the myocardium (7). Catalase potentially offers an advantage over the glutathione peroxidase system in the ischemic myocardium in that its ability to break down H2O2 is not sensitive to the intracellular redox state. Based on the above assertions, one would predict that increasing catalase levels in the heart would enhance the defense against hydrogen peroxide during periods of ischemia and reperfusion. The administration of exogenous antioxidant enzymes, including catalase, for cardiac protection has been studied in animals and humans. While variable protection was apparent in some studies (9, 10), others showed no protective benefit (11, 12). These inconsistent results may be explained by the instability of exogenously administered antioxidant enzymes and the failure of cardiocyte uptake, rather than arguing against free radicals as the primary mediator of ischemia–reperfusion injury. Myocardial gene transfer may be useful in assessing the relative contribution of free radicals in ischemia–reperfusion injury. High levels of catalase can be achieved by constitutive expression of catalase cDNA transfected into the myocardium via replication-deficient adenovirus. In this study, we examined whether the gene transfer of catalase can provide protection against ischemia–reperfusion injury in the hearts of rabbits. MOLECULAR THERAPY Vol. 2, No. 5, November 2000 Copyright  The American Society of Gene Therapy 1525-0016/00 $35.00

ARTICLE MATERIALS

AND

METHODS

Animal care. This study was performed in accordance with the animal care guidelines of the Institutional Animal Care and Use Committee of the University of Pennsylvania, which conform to the policies of the American Heart Association and the current federal guidelines. Adenoviral vector construction. Replication-deficient adenoviral vectors containing the transgene encoding β-galactosidase (LacZ) or human catalase (CAT), driven by the constitutively active cytomegalovirus promoter (CMV), were generated by homologous recombination with the human serotype 5 adenoviral DNA in human embryonic kidney stem cells (13). The Ad.CMV.LacZ and Ad.CMV.CAT vectors were then propagated in 293 cells, purified in CsCl density purification, dialyzed, and stored at −70C until use. The titer of each viral preparation was determined by plaqueforming assay in 293 cells. In vivo cardiac gene delivery by direct myocardial injections. New Zealand White male rabbits (3–4 kg, 10–16 weeks of age) were anesthetized with ketamine (50 mg/kg), xylazine (6 mg/kg), and acepromazine (1 mg/kg) via intramuscular injection. The rabbit was intubated and mechanically ventilated (Model 683, Harvard Apparatus, North Billerica, MA) to maintain arterial PO2 above 80 mmHg and pH of 7.35 to 7.45. A left thoracotomy was performed through the fourth intercostal space, the heart was exposed, and the circumflex artery was identified. A total volume of 400 µl of adenovirus (5 × 1011 pfu/ml) encoding for LacZ or catalase was delivered into the circumflex distribution in five equal aliquots (14) (Fig. 1). The left thoracotomy was closed in layers and the rabbit was extubated and allowed to recover. Assay for β-galactosidase expression. Three days after injection of Ad.CMV.LacZ, rabbits were euthanized and their hearts and livers were harvested. Residual blood in the hearts was flushed out with Dulbecco’s phosphate-buffered saline (DPBS, 1 mM KH2PO4, 10 mM Na2HPO4, 137 mM NaCl, 2.7 mM KCl, pH 7.4). The hearts and livers were bathed in freezing compound (TissueTek OCT compound, pH 7.4), embedded in 50-ml tubes, and snap-frozen in liquid nitrogen. Cross-sections of the heart and liver in 14-µm thickness at 100-µm intervals were mounted on microscopic slides. The entire heart, from the base to the apex, was sectioned. At room temperature, the heart and liver sections were fixed separately for 10 min in 4% glutaraldehyde and 0.2% paraformaldehyde solution in DPBS. These sections were then rinsed three times in DPBS and incubated at 37C for 4 h in X-gal solution containing 2 mmol/L 5-

FIG. 1. In vivo rabbit heart preparation for direct myocardial injection of Ad.CMV.LacZ or Ad.CMV.CAT into LV supplied by the lateral branch of the left coronary artery. Multiple injections were used to deliver 4 × 1011 particles of adenoviral vectors (400 µl in volume) into a 1.5 × 1.5-cm area of LV, shown as the dark circular area.

MOLECULAR THERAPY Vol. 2, No. 5, November 2000 Copyright  The American Society of Gene Therapy

FIG. 2. In vivo rabbit heart preparation for measurement of LV segment shortening of the region 3 days after injection of the adenoviral vectors. Two 1-mm piezoelectric sonomicrometry crystals were implanted in the subendocardium approximately 8 mm apart in the injected LV area.

bromo-4-chloro-3-indolyl-D-galactopyranoside, 4 mmol/L potassium ferricyanide, 4 mmol/L potassium ferrocyanide, and 1 mmol/L MgCl2 in DPBS. The X-gal reaction was terminated by washing the sections three times with DPBS. The sections were then fixed with 4% glutaraldehyde and 0.2% paraformaldehyde solution in DPBS at 4C, followed by washing with DPBS three times and counterstaining with hematoxylin and eosin. The specimens were examined using light microscopy to determine the extent of β-galactosidase expression. Spectrophotometric assay of catalase enzyme activity. Three days after injection of Ad.CMV.LacZ (LacZ, n = 5) or Ad.CMV.CAT (CAT, n = 5), rabbits were euthanized and their hearts were explanted and perfused via the ascending aorta with DPBS at 4C to flush out residual blood in the coronary circulation. The area of the left ventricular (LV) free wall injected with the vectors was harvested for catalase activity assay. A similar area of the interventricular septum (septum) not injected with the vectors was also harvested as control. An additional group of rabbits (native, n = 5) not subject to viral injection were used for comparison. Heart tissues were homogenized with 4 vol of ice-cold 10 mmol/L Tris–HCl (pH 7.5) buffer containing 0.25 mol/L sucrose, 0.5 mmol/L DTT, 1 mmol/L EDTA, and 0.1 mmol/L PMSF. The homogenate was sonicated on ice for 30 s followed by centrifugation at 700g for 15 min at 4C. Protein concentrations were determined by Bio-Rad protein assay (BioRad, Hercules, CA). Catalase activity in the postnuclear supernatant was assayed by hydrogen peroxide degradation assay (15) where 25 µg of specimen protein was incubated with 30 µl of 1% H2O2 in 1 ml of 50 mmol/L phosphate/0.02% Triton X-100 buffer. The disappearance of H2O2 was followed spectrophotometrically at 240 nm. One unit of enzyme activity is defined as the decomposition of 1 µmol of H2O2 per minute at 25C and pH 7.0. Western blot analysis of catalase enzyme expression. Native rabbits, rabbits 3 days after injection of Ad.CMV.LacZ, or rabbits 3 to 14 days after injection of Ad.CMV.CAT were euthanized. Their hearts were harvested and prepared as described in the previous section. One hundred micrograms of proteins from each sample was electrophoresed on 12% SDS–polyacrylamide gel and transferred to nitrocellulose sheet (16). The blot was immunoprobed with sheep anticatalase (Binding Site, Birmingham, UK) and the bound alkaline phosphatase-conjugated second antibody was detected following a color reaction. Assessment of regional LV function during ischemia and reperfusion. Regional LV function was assessed by the technique of crystal sonomi-

471

ARTICLE crometry. Two 1-mm piezoelectric sonomicrometry crystals (Triton Technology, San Diego, CA) were implanted in the myocardium supplied by the circumflex artery (Fig. 2). LV segment shortening was measured by the distance between the two crystals during each cardiac cycle. Fractional systolic shortening (FSS), a unitless measure of myocardial contractility, was defined by the equation FSS = (end diastolic length − end systolic length) ÷ (end diastolic length), where end diastolic length was measured at the onset of the left ventricular isovolumic contraction and end systolic length was measured at the maximum negative rate of pressure rise (dP/dt). Ten consecutive heartbeats were analyzed at each time point of the experiment. Three days after injection of 4 × 1011 particles of Ad.CMV.LacZ (LacZ, n = 5) or 4 × 1011 particles of Ad.CMV.CAT (CAT, n = 5), the rabbits were reintubated and mechanically ventilated to maintain pH between 7.35 and 7.45 and pO2 greater than 80 mmHg. Animals were monitored with arterial line, EKG, and pulse oxymetry. Intravenous lidocaine infusion was started 15 min prior to surgical manipulation and continued throughout the experiment. This was done in order to prevent the development of arrhythmias during reperfusion. In addition to the LacZ and CAT groups, five rabbits (native) were used for comparison. Through a repeated left thoracotomy, the hearts were exposed and the areas previously injected with Ad.CMV.LacZ or Ad.CMV.CAT were identified. Two 1mm sonomicrometry crystals were implanted into the myocardium approximately 8 mm apart, on either side of the coronary artery. In the native group, a similar region of the LV was chosen for placement of the crystals. The hearts were allowed to equilibrate for 15 min. A 5-O silk suture was passed around the coronary artery proximal to the injected area, and the ends of the silk sutures were threaded through a small vinyl tube (Fig. 2). After acquiring baseline data, the coronary artery was occluded by pulling the snare, which was then secured with a mosquito clamp. Myocardial ischemia was confirmed by ST-segment elevation on the EKG and cyanotic appearance of the ischemic region. For reversible ischemia–reperfusion injury, the coronary artery was occluded for 15 min and reperfused by releasing the snare. Reperfusion was confirmed by color change on the myocardial surface and the gradual return of the ST segment to the baseline. All hearts were allowed to reperfuse for 2 h. Sonomicrometry signals were recorded at baseline prior to occlusion, 15 min after occlusion, and 2, 10, 30, 60, 75, 90, 105, and 120 min during reperfusion. FSS was calculated for each animal at each

time point of the experiment. Comparison was made among the three treatment groups: native (n = 5), LacZ (n = 5), and CAT (n = 5). After the functional study was concluded, the suture was again fastened and methylene blue dye was injected into the left atrium. The area not stained was taken as the area-at-risk and procured for histological evaluation. TUNEL assay. Specimens were obtained from storage and allowed to warm to 22C. Five transverse 10-µm sections were prepared with a cryostat at 0.25-cm intervals from the point of circumflex occlusion to the apex of the heart. TUNEL assay was performed with a TdT-FragEL DNA fragmentation detection kit (Oncogene Research Products). Statistical methods. Statistical analysis was performed with a commercially available statistical software package (SigmaStat, Jandel Scientific Software). Unpaired, two-tailed Student’s t tests were used to compare catalase enzyme activity and FSS between the treatment groups. The results were expressed as means ± SEM. A difference was considered statistically significant if P < 0.05.

RESULTS Efficiency of Gene Transfer Efficiency of our direct adenoviral injection approach was evaluated by X-gal reaction to detect β-galactosidase expression. At 3 days after injection of 5 × 1011 particles of Ad.CMV.LacZ (Fig. 3), there was transmural expression of β-galactosidase (stained blue) in the injected region of the LV. On examination with higher magnification, gene transfer efficiency was approximately 40–60% of the injected area. Rarely was mechanical destruction of myocytes seen as a result of the injection. Inflammatory response was minimal at 3 days following injection as evidenced by the paucity of mononuclear cells seen in the injected region. No β-galactosidase expression was seen in the liver (data not shown). Direct in vivo gene

FIG. 3. Histologic examination of β-galactosidase gene expression in rabbit hearts injected with Ad.CMV.LacZ. The blue-staining area denotes positive β-galactosidase expression. There was transmural expression of β-galactosidase in the LV injected with Ad.CMV.LacZ.

472

MOLECULAR THERAPY Vol. 2, No. 5, November 2000 Copyright  The American Society of Gene Therapy

ARTICLE TABLE 1 Results of the Spectrophotometric Assay for Catalase Enzyme Activity

FIG. 4. Catalase enzyme activities as determined by spectrophotometric assays. There was significant elevation of catalase enzyme activity in CAT LV alone and a 4.27-fold (P = 0.0012), 5.69-fold (P < 0.001), and 4.23-fold (P < 0.001) increase compared with native LV, LacZ LV, and CAT septum, respectively.

transfer by multiple myocardial injections provided an efficient means of local gene delivery in the heart, as demonstrated by the expression of the reporter protein beyond the immediate vicinity of the needle injection site.

Enhanced Catalase Activity To determine the amount of increased catalase activity that resulted from viral gene transfer, spectrophotometric assay of catalase activity was performed on heart tissues 3 days after injection with 4 × 1011 particles of Ad.CMV.CAT (CAT LV, n = 5). As shown in Table 1 and Fig. 4, catalase enzyme activity was significantly elevated only in the LV region injected with Ad.CMV.CAT (14.6 ± 1.6 U/mg protein). This represented statistically significant 4.3-fold (P = 0.0012), 5.7-fold (P < 0.001), and 4.2fold (P < 0.001) increases in catalase activity compared to tissues from the native LV, LacZ LV, and CAT septum, respectively. These results demonstrate that adenoviral gene transfer via direct injection yielded a significantly elevated level of functionally active enzyme.

Catalase activity in uninjected septum (U/mg protein)

Catalase activity in left ventricle (U/mg protein)

Native

3.4  0.2

3.6  0.3

LacZ

2.6  0.4

2.9  0.5

CAT

3.5  0.6

14.6  1.6

Note. Expressed as activity units/mg protein  SEM. The native group (n = 5) received 11 no virus, the LacZ group (n = 5) received 4  10 particles of Ad.CMV.LacZ, and the CAT grop (n = 5) received 4  1011 particles of Ad.CMV.CAT. Note that only the region of the left ventricle receiving the viral injection was assayed and that the septum was uninjected in all cases.

Ad.CMV.CAT, which correlated with the increased catalase enzyme activities demonstrated by the spectrophotometric assay.

Assessment of Regional Myocardial Function during Reversible Ischemia and Reperfusion Fifteen rabbits were divided into three groups: rabbits with no viral injection (native, n = 5), rabbits 3 days after injection of 4 × 1011 particles of Ad.CMV.LacZ (LacZ, n = 5), and rabbits 3 days after injection of 4 × 1011 particles of Ad.CMV.CAT (CAT, n = 5). FSS was calculated at each time point of the experiment and graphed in Fig. 6. Baseline heart rates and arterial blood pressures were similar among the three experimental groups. At baseline, FSS ranged from 16 to 18.8% and was not significantly different among the three groups. At 15 min of occlusion, FSS was similar among the three groups: 6.03, 7.34, and 7.38%, with decreases of 62.3, 56.6, and 60.7% from baseline in the native, LacZ, and CAT hearts, respectively. Perfusion was restored after 15 min of coronary occlusion. There was a similar increase in FSS at 2

Western Blot Analysis of Catalase Enzyme Expression To confirm that the enhanced antioxidant activities in the LV injected with Ad.CMV.CAT were caused by increased catalase expression, Western blot analysis was performed on tissues harvested from native, LacZ, and CAT hearts. Representative Western blot samples are shown in Fig. 5. There was a significantly increased catalase protein expression in Ad.CMV.CAT-injected LV at 3 days following injection (lanes 5 and 6) compared to CAT septum (lane 4), LacZ LV (lane 3), native LV (lane 1), and native septum (lane 2). The significantly increased catalase expression persisted at 7 days after injection (lane 7). At 14 days after injection (lanes 8 and 9), catalase level was still higher than that of controls, but was lower than those at 3 and 7 days after injection. The Western blot results demonstrated that catalase expression was enhanced in the myocardium injected with MOLECULAR THERAPY Vol. 2, No. 5, November 2000 Copyright  The American Society of Gene Therapy

FIG. 5. Western blot analysis of catalase protein expression in the LV region injected with 4 × 1011 particles of Ad.CMV.CAT (CAT) or Ad.CMV.LacZ (LacZ) and LV not subjected to adenoviral injection (native). The interventricular septum was also harvested for Western blot analysis. Equal amounts of proteins were loaded in each lane. Lane 1, native LV; lane 2, native septum; lane 3, LacZ LV 3 days after injection; lane 4, CAT septum 3 days after injection; lanes 5 and 6, CAT LV 3 days after injection; lane 7, CAT LV 7 days after injection; and lanes 8 and 9, CAT LV 14 days after injection. Significantly increased catalase expression in lanes 5, 6, and 7, moderately increased catalase expression in lanes 8 and 9, and baseline low catalase expression in lanes 1 to 4 (controls) were demonstrated.

473

ARTICLE sion using adenoviral gene transfer. Our longest period of expression was 4 weeks. By Western blots, the expression was markedly diminished by 3 weeks, and no expression was seen 1 month following injection of virus. However, we did not see a diminution on contractile function at the 4-week time point (data not shown).

DISCUSSION

FIG. 6. FSS in three treatment groups undergoing 15 min of regional ischemia followed by 2 h of reperfusion. The three treatment groups were rabbit hearts after 3 days of injection of Ad.CMV.CAT (l, CAT; n = 5) or Ad.CMV.LacZ (n, LacZ; n = 5) and hearts with no injection (s, native; n = 5). Significantly decreased FSS occurred in all groups after 15 min of occlusion. FSS was preserved in CAT rabbit hearts following reperfusion. Both native and LacZ hearts showed similar injury patterns following reperfusion with recovery of contractile function after 2 h of reperfusion.

min following reperfusion. This initial hypercontractility appeared to be due to realkalinization and sudden release of intracellular calcium upon reperfusion (17). At 10 min after reperfusion, FSS of the CAT hearts was 17.5%, significantly different from those of the native (7.41%, P = 0.006) or LacZ (8.46%, P = 0.0107) hearts. The FSS was preserved in the CAT hearts during reperfusion, ranging from 17.5 to 20.4%, not significantly different from the baseline of 18.8%. In contrast, the native and LacZ hearts showed a significant decrease in FSS at 10 min of reperfusion. Subsequent recovery of contractile function was demonstrated in the native and LacZ hearts, where the FSS was 15.6 and 15.2% at 2 h of reperfusion. These results showed that full recovery of regional myocardial function was possible after 15 min of coronary occlusion. The native and LacZ hearts showed virtually identical patterns of reperfusion injury. There was no statistically significant difference in FSS at any time point of the experiment. This demonstrated that the method of gene delivery by direct myocardial injection did not cause preferential or detrimental effects on myocardial function. The observed benefit in hearts treated with Ad.CMV.CAT can only be attributed to the increased intracellular catalase activities. We examined the area in risk in all groups for histological evidence of apoptosis (using TUNEL assays). However, with the 15-min ischemia–reperfusion protocol used in this study, no TUNEL-positive cells were seen in any sections from any of the experimental groups. While the long-term effects of catalase overexpression would be of interest, we achieved only transient expres-

474

Considerable experimental evidence suggests that reactive oxygen species are generated during reperfusion of the ischemic myocardium and may contribute to myocardial ischemia–reperfusion injury. Under normal physiologic conditions, the myocardium is equipped with the glutathione peroxidase system, which efficiently metabolizes these reactive species. However, when subjected to ischemia and reperfusion, there is excessive production of free radicals that may overwhelm the antioxidant defense mechanism, combined with a disadvantageous intracellular redox stage that prevents reconstitution of the system. Both catalase and glutathione peroxidase can metabolize H2O2. Under normal conditions, the low rate of H2O2 production is preferentially cleared by glutathione peroxidase (18). Glutathione peroxidase catalyzes the conversion of H2O2 to water by oxidizing GSH to GSSG. The GSH/GSSG ratios in normal cells are kept high by the enzymatic reaction of glutathione reductase, which converts GSSG to GSH in the presence of NADPH as a reducing agent. In the absence of an adequate quantity of NADPH, as in ongoing ischemia, the GSH/GSSG ratio falls, disabling the glutathione peroxidase (5). Reperfusion further compounds the problem by excessive production of H2O2 overwhelming the capacity of regenerated GSH (8). This concept is supported by studies demonstrating that the recovery of contractile function following 30 min of ischemia is concomitant with a complete restoration of the GSH/GSSG ratio (19, 20). Although catalase levels are low in the myocardium, studies have shown that it participates to a significant extent in the detoxification of H2O2 (7) during periods of ischemia/reperfusion. Initial attempts were made to increase antioxidant supply by loading the heart by intravenous administration or by addition to the pump perfusate during cardiopulmonary bypass. Exogenous antioxidant enzyme administration, however, is limited by the ability of the myocytes to internalize these large molecules. Strategies to increase intracellular catalase activity, such as administration of endotoxin and subjecting animals to heat stress, have been shown to improve contractile function during reperfusion (21, 22). In this study we demonstrated that regional adenoviralmediated gene transfer could be effectively used to significantly increase the levels of catalase activity in the myocardium. This resulted in the preservation of contractile function (elimination of stunning) during reperfusion in the transfected region. It is likely that the increased catalase enzyme activity enabled the heart to MOLECULAR THERAPY Vol. 2, No. 5, November 2000 Copyright  The American Society of Gene Therapy

ARTICLE more effectively metabolize the overproduction of H2O2 that resulted from a combination of increased free radical production and decreased glutathione peroxidase function. Thus, this study supports the hypothesis that increased catalase expression ameliorates myocardial contractile dysfunction from ischemia–reperfusion injury and suggests that inability to rapidly degrade H2O2 is the primary cause of reversible ischemia–reperfusion injury. Gene transfer of catalase into the heart may someday provide a means to protect against transient bouts of myocardial ischemia. ACKNOWLEDGMENTS This work was supported by grants from the NIA (AG133329) and NHLBI (HL07843).

REFERENCES 1

Weisfeldt, M. L. (1987). Reperfusion and reperfusion injury. Clin. Res. 35: 13–20. Becker, L. C., and Ambrosio, G. (1987). Myocardial consequences of reperfusion. Prog. Cardiovasc. Dis. 30: 23–44. 3 Bolli, R., et al. (1989). Direct evidence that oxygen-derived free radicals contribute to postischemic myocardial dysfunction in the intact dog. Proc. Natl. Acad. Sci. USA 86: 4695–4699. 4 Gardner, T. J. (1994). Oxygen radicals and myocardial stunning. J. Card. Surg. 9(Suppl.): 422–444. 5 Ferreri, R., et al. (1985). Oxygen-mediated myocardial damage during ischemia and reperfusion: Role of the cellular defenses against oxygen toxicity. J. Mol. Cell. Cardiol. 17: 937–945. 6 Singh, A., Lee, K. J., Lee, C. Y., Goldfarb, R. D., and Tsan, M. F. (1989).Relation between myocardial glutathione content and extent of ischemia–reperfusion injury. Circulation 80: 1795–1806. 7 Steare, S. E., and Yellon, D. M. (1995). The potential for endogenous myocardial antioxidants to protect the myocardium against ischemia–reperfusion injury: Refreshing the parts exogenous antioxidants cannot reach? J. Mol. Cell. Cardiol. 27: 65–74. 2

MOLECULAR THERAPY Vol. 2, No. 5, November 2000 Copyright  The American Society of Gene Therapy

8

Curello, S., Ceconi, C., Bignoli, C., Ferrari, R., Albertini, A., and Guarnieri, C. (1985). Changes in cardiac glutathione status after ischemia and reperfusion. Experientia 41: 42–43. 9 Johnson, D. L., Horneffer, P. J., Dinatale, J. M., Gott, V. L., and Gardner, T. J. (1987). Free radical scavengers improve functional recovery of stunned myocardium in a model of surgical coronary revascularization. Surgery 102: 334–340. 10 Jeroudi, M. O., Triana, F. J., Patel, B. S., and Bolli, R. (1990). Effect of superoxide dismutase and catalase given separately, on myocardial “stunning.” Am. J. Physiol. 259: H889–H901. 11 Bjorkman, J. A., Sutherland, I., Gustafsson, D., Sjoquist, P. O., and Abrahamsson, T. (1991). Superoxide dismutase and catalase do not improve recovery of regional myocardial contractile function when given at the time of reperfusion after reversible regional ischemia in anesthetized dogs. Basic Res. Cardiol. 86: 236–244. 12 Flaherty, J. T., et al. (1994). Recombinant human superoxide dismutase (h-SOD) fails to improve recovery of ventricular function in patients undergoing coronary angioplasty for acute myocardial infarction. Circulation 89: 1982–1991. 13 Kozarsky, K. F., and Wilson, J. M. (1993). Gene therapy: Adenoviral vectors. Curr. Opin. Genet. Dev. 3: 499–503. 14 Podesser, B., et al. (1997). Epicardial branches of the coronary arteries and their distribution in the rabbit heart: The rabbit heart as a model of regional ischemia. Anat. Rec. 247: 521–527. 15 Aebi, H. (1984). Catalase in vitro. Methods Enzymol. 105: 121–126. 16 Towbin, J., Staehelin, T., and Gordon, J. (1979). Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: Procedure and some applications. Proc. Natl. Acad. Sci. USA 76: 4350–4354. 17 Ward, C. A., and Moffat, M. (1995). Modulation of sodium–hydrogen exchange activity in cardiac myocytes during acidosis and realkalinization: Effects on calcium, pH and cell shortening. Cardiovasc. Res. 29: 247–253. 18 Halliwell, B., and Gutteridge, J. M. C. (1989). Free Radicals in Biology and Medicine, 2nd ed. Clarendon Press, Oxford. 19 Halliwell, B., and Gutteridge, J. M. C. (1984). Oxygen toxicity, oxygen radicals, transition metals and disease. Biochem. J. 219: 1–14. 20 Ferrari, R., Curello, S., Ceconi, C., Cargnoni, A., Condorelli, E., and Albertini, A. (1988). Alterations of glutathione status during myocardial ischemia and reperfusion. In Oxygen Radicals in the Pathophysiology of Heart Disease (P. K. Singal, Ed.), pp. 145–161. Kluwer Academic, Dordrecht/Norwell, WA. 21 Brown, J. M., et al. (1989). Endotoxin pretreatment increases endogenous catalase activity and decreases ischemia–reperfusion injury of isolated rat hearts. Proc. Natl. Acad. Sci. USA 86: 2516–2520. 22 Mocanu, M. M., Steare, S. E., Evans, M. C. W., Nugent, J. H., and Yellon, D. M. (1993). Heat stress attenuates free radical release in isolated perfused rat heart. Free Radical Biol. Med. 15: 459–463.

475