Does Tumor Necrosis Factor-α (TNF-α) Contribute to Myocardial Reperfusion Injury in Anesthetized Rats?

Gen. Pharmac. Vol. 32, No. 1, pp. 41–45, 1999 Copyright  1998 Elsevier Science Inc. Printed in the USA.

ISSN 0306-3623/99 $–see front matter PII S0306-3623(98)00050-0 All rights reserved

Does Tumor Necrosis Factor-a (TNF-a) Contribute to Myocardial Reperfusion Injury in Anesthetized Rats? Matthew McVey, Mark H. Perrone and Kenneth L. Clark* Department of Cardiovascular Discovery (NW4), Rhoˆne-Poulenc Rorer, 500 Arcola Road, Collegeville, PA 19426, USA [Tel: 610-454-5101; Fax: 610-454-8740; E-mail: [email protected]] ABSTRACT. 1. This study examined the potential role of tumor necrosis factor-alpha (TNF-a) in myocardial ischemia-reperfusion injury using an anesthetized rat model of myocardial infarction. 2. The key endpoints were infarct size and serum TNF-a levels (measured by a specific ELISA technique). 3. Three groups of rats were studied: vehicle controls (n56); positive controls for infarct size reduction (ischemic preconditioning; n56); and a group treated with the selective inhibitor of PDE-IV and TNF-a production, rolipram (1mg/kg IV 10-min prereperfusion11 mg/kg per minute through 1-hr reperfusion, n56). 4. Baseline preischemia levels of serum TNF-a were low (z0.1 nM) and showed a trend for further reduction in all treatment groups at 1 min and 3 hr into the postischemia reperfusion period. 5. Infarct size (6862% of the ischemic area in controls) was significantly reduced (41% decrease) by preconditioning, but was unchanged in rolipram-treated animals. 6. Collectively, these data argue against an important role for TNF-a in lethal reperfusion injury in this rat model of myocardial infarction. gen pharmac 32;1:41–45, 1999.  1998 Elsevier Science Inc. KEY WORDS. TNF-a, myocardial infarction, anesthetized rat, PDE-IV inhibition, rolipram, ischemic preconditioning

INTRODUCTION The use of strategies to reopen occluded coronary arteries has markedly advanced the treatment of acute myocardial infarction. Nevertheless, substantial (more than 10%) loss of viable ventricular tissue occurs in a significant number of patients, and it is established that infarct size is closely related to subsequent morbidity and mortality (Miller et al., 1995). Thus, there remains an unmet medical need for agents that can be given in conjunction with thrombolytic therapy or angioplasty to further reduce infarct size. The discovery of such novel drugs first requires an understanding of the processes leading to myocardial injury during myocardial ischemia-reperfusion. Although reperfusion of ischemic myocardium is essential to save tissue, the controversial hypothesis exists that, paradoxically, reperfusion may also trigger an inflammatory process in the myocardium, which actually causes a degree of lethal cell injury. (Granger and Korthuis, 1995; Hansen, 1995; Hearse and Bolli, 1991). This has prompted many investigators to study changes in the levels of proinflammatory cytokines, which occur upon reperfusion of ischemic myocardium. In that regard, increases in cardiac and systemic levels of tumor necrosis factor-alpha (TNF-a) have been reported in both preclinical and clinical studies (Gurevitch et al., 1996; Herskowitz et al., 1995; Jansen et al., 1992; Maury and Teppo, 1989; Squadrito et al., 1993). This study aimed to further evaluate the role of TNF-a in myocardial ischemia-reperfusion injury. Using a rat model of myocardial in*To whom correspondence should be addressed. Received 14 October 1997.

farction two approaches were taken. First, systemic levels of TNF-a were measured before and after ischemia and reperfusion using an immunoassay (ELISA) specific for rat TNF-a. Second, because inhibition of phosphodiesterase type IV (PDE-IV) has been shown to reduce TNF-a production (Bergman and Holycross, 1996; Prabhakar et al., 1994; Semmler et al., 1993), we assessed whether administration of the selective PDE-IV inhibitor rolipram (Reeves et al., 1987; Shahid and Nicholson, 1990), just prior to reperfusion, could reduce myocardial infarct size. MATERIALS AND METHODS All experiments were conducted in accordance with a protocol approved by the Rhoˆne-Poulenc Rorer Animal Care and Use Committee and conform to the NIH Guidelines for the Use and Care of Laboratory Animals. Male Sprague–Dawley rats (434 to 558 g), anesthetized with 120 mg/kg IP, thiobutabarbital, (Inactin, BYK-Gu¨lden), were used. A tracheotomy was performed to ensure airway patency and to allow for mechanical positive pressure ventilation. A catheter was inserted into the right common carotid artery and connected to a pressure transducer for the measurement of arterial pressure. The right external jugular vein was cannulated for the administration of test compounds. The animal was ventilated using 0.7 ml of room air per 100 g of body weight at 55 to 58 cycles/min. The heart was exposed via an intercostal approach. The pericardium was gently dissected. Once the location of the coronary artery was ascertained, a 4-0 silk suture was placed around the vessel. This was accomplished by plunging a taper needle attached to the suture through the myocardium and around the vessel location. Electrocar-

42

M. McVey et al.

diogram (ECG) needle electrodes were then placed to monitor Lead II ECG changes. Arterial pressure and ECG were recorded on a Modular Instruments MI2 data processing system. Computerized data acquisition from the pressure and ECG traces yielded the following measured parameters: mean arterial pressure, heart rate and rate-pressure product. The animals were allowed to stabilize for an hour to allow for recovery from the surgical preparation.

Experimental protocol Animals were divided into three groups of 6 rats each; control; rolipram-treated; and ischemic preconditioning. Preconditioning was achieved using three series of 3-min coronary occlusion followed by 5 min of reperfusion. Myocardial infarction was induced with a 30min coronary artery occlusion, followed by a 3-hr reperfusion period. Rolipram-treated rats received a 1-mg per kg bolus followed by a 10mg/kg per minute infusion starting 10 min prior to and continuing through 1 hr of reperfusion. Control rats were infused with vehicle (2.5% DMSO, 23.75% PEG-200, 73.75% normal saline) over the same 70-min period. The dose of rolipram used was based on literature reports by other workers who administered the compound intravenously to rats and dogs and demonstrated pharmacological activity associated with PDE-IV inhibition (Chiu et al., 1996; Prabhakar et al., 1994; Simpson et al., 1992). Arterial blood samples of 0.5 ml were drawn at 45 min of baseline stabilization, 1 min of reperfusion and after 3 hr of reperfusion. Blood samples were kept on ice for 1 hr at which point they were centrifuged. Serum was then removed and kept frozen until required for analysis. Quantitative analysis was made using an ELISA specific for rat TNF-a with a sensitivity limit of 1 pg/ml (Biosource International). At the end of the 3 hr of reperfusion, the coronary artery was reoccluded. Patent blue dye, at 30 mg/kg, was injected into the apex of the left ventricle to stain the region of the heart not perfused by the occluded vessel. The heart was stopped with intravenous KCl and rapidly excised. After the atria and right ventricle were removed, the left ventricle was quick frozen at 2708C for 10 min. It then was sliced into five sections from apex to base (occlusion point). These rings were placed into a Petri dish and submerged with cold saline. The area at risk (i.e., the area of tissue supplied by the occluded artery) was then determined using a computerized image system. After the areas at risk were determined, the slices were transferred to another dish containing warm (378C) 0.25% nitroblue tetrazolium (NTB). NTB stains healthy myocardium purple, whereas infarcted areas appear paler and hemorrhaged tissue does not stain. After 20 to 25 min, the infarcted tissue was visible and the area of necrosis was measured. The weight of each slice was then determined and the overall infarct and risk proportions were calculated. Data are expressed as mean6SEM. Statistical significance was assessed using ANOVA followed by Dunnett’s posthoc test to compare treatment groups with control. Statistical significance was assigned at P,0.05.

FIGURE 1. Effects of rolipram or preconditioning on area at risk as a percentage of left ventricle (AAR/LV), infarct size as a percentage of area at risk (IS/AAR) and serum TNF-a levels in anesthetized rats subjected to myocardial ischemia-reperfusion. All groups: n56 (mean6SEM). Baseline values for TNF-a515 min prior to ischemia. *P,0.05, ANOVA/Dunnett’s test.

Baseline levels of serum TNF-a were measured 15 min prior to ischemia and were low (z0.1 nM) in all groups. One minute into the reperfusion period, TNF-a levels remained unchanged relative to baseline measurements, although there was a trend toward a reduction in the preconditioned group. By 3 hr into the reperfusion period, reductions in TNF-a levels were observed in all treatment groups.

RESULTS

Infarct size and serum TNF-a levels

Cardiovascular function

Myocardial ischemia (30 min) followed by reperfusion (3 h) produced an area at risk of ischemia (35% to 40% of the left ventricle), which was similar in all treatment groups. In control animals, 69% of the area at risk became infarcted, which was almost identical to lesions (71% of area at risk) observed in animals that received the PDE-IV inhibitor, rolipram, 10 min prior to reperfusion (1 mg/kg IV110 mg/kg per minute through hour 1 of reperfusion). In contrast to rolipram, ischemic preconditioning significantly reduced infarct size by 41% (Fig. 1).

Changes in blood pressure, heart rate and rate pressure product in the different treatment groups are shown in Figure 2. Blood pressure tended to drop in all groups during the latter half of the reperfusion period, whereas heart rate was not changed. Final readings of blood pressure, heart rate and rate pressure product were similar in all groups. When administered during the ischemic period 10 min prior to reperfusion, rolipram induced a significant fall in pressure of approximately 25 mm Hg, which was not observed in the vehicle group. Consequently, rolipram caused a modest but significant decrease in rate

TNF-a and Myocardial Infarction

FIGURE 2. Effects of rolipram (n), preconditioning (.) or vehicle (j) on blood pressure, heart rate and rate pressure product in anesthetized rats subjected to myocardial ischemia-reperfusion (I5ischemia, R5reperfusion). All groups: n56, (mean6SEM). P,0.05, ANOVA/Dunnett’s test.

pressure product, indicative of decreased cardiac afterload and oxygen demand. DISCUSSION Although the treatment of myocardial infarction has markedly improved in the past decade, significant cardiac necrosis still occurs in many patients, and it is established that final infarct size closely cor-

43 relates with subsequent morbidity and mortality (Miller et al., 1995). Thus, it is important to understand the factors and pathways that contribute to infarct development. In this regard, many investigators have focused on the controversial concept of lethal reperfusion injury (Granger and Korthuis, 1995; Hansen, 1995; Hearse and Bolli, 1991), which suggests that, although reperfusion of ischemic myocardium is essential for tissue salvage, it also paradoxically triggers inflammatory processes that can, in their own right, cause a degree of myocyte loss. In that regard, it is of interest to note that several groups (Gurevitch et al., 1996; Herskowitz et al., 1995; Jansen et al., 1992; Squadrito et al., 1993) have reported an increase in the expression or production of the proinflammatory cytokine TNF-a following myocardial ischemia-reperfusion and suggested that it may contribute to infarct development. Thus, the aim of the current study was to further probe the role of TNF-a in myocardial ischemia-reperfusion injury. Two main experimental approaches were taken. First, a specific ELISA technique was used to measure TNF-a levels prior to myocardial ischemia and then at two further timepoints during the phase of myocardial reperfusion. Second, we determined whether administration of a selective PDE-IV inhibitor rolipram, just prior to reperfusion and through the first hour of the reperfusion period, could reduce infarct size. The rationale for the second part of the study was based on the observation that PDE-IV inhibition is known to be a highly effective mechanism for inhibition of TNF-a. In that regard, rats were a suitable species of choice for these experiments as PDE-IV is the principal phosphodiesterase enzyme present in rat heart (Shahid and Nicholson, 1990) and PDE-IV inhibition has been shown to inhibit TNF-a production by rat myocardial tissue (Bergman and Holycross, 1996). Ischemic preconditioning triggers endogenous protective mechanisms in the myocardium and was used as a positive control to demonstrate the sensitivity of the model to infarct size reduction. Ischemic preconditioning was first demonstrated by Murry et al. (1986) and has since become accepted as an effective procedure to reduce infarct size in all species studied to date. Results from the first part of the study indicate that the model used was robust and reproducible in terms of the area of myocardium that was made ischemic (z40% of the left ventricle in all treatment groups). In control animals, 30-min ischemia and 3 hr reperfusion produced a large and reproducible infarct (z69% of the ischemic area). Measurements of serum TNF-a levels prior to and postischemia yielded similar data in all treatment groups (control, rolipramtreated and preconditioned) and will therefore be discussed collectively. Thus, baseline preischemia serum concentrations of TNF-a were detectable but low (z0.1 nM). It is worth noting that the Kd of TNF-a for its receptor was found to be in the range 0.3 to 1 nM (Pang et al., 1989; Van Bladel et al., 1991), suggesting that the levels of TNF-a present at baseline were unlikely to have been exerting significant physiological effects. Postischemia, 1 min into the reperfusion period, no increase in TNF-a levels was observed. Indeed, a nonsignificant trend for a reduction in TNF-a levels was observed in the rolipram-treated and preconditioned animals. By 3 hr into the reperfusion period, marked decreases in TNF-a levels were observed in all treatment groups, although statistical significance was not achieved in the control group due to baseline variability. Why TNF-a levels were higher at baseline that at the end of the reperfusion period is not clear, although it is possible that the anesthesia or surgical stress cause an increase above normal background levels. However, as discussed earlier, baseline levels are modest and the important observation is that only decreases are observed following ischemia and reperfusion. Therefore, the present data contrast somewhat with previous reports of increased levels of TNF-a following reperfusion of rat ische-

44 mic myocardium (Gurevitch et al., 1996; Squadrito et al., 1993) although they are in agreement with previous observations in a canine model (Field et al., 1994). Why the discrepancy with previous rat data? It could be argued that, in the present study, by only measuring TNF-a levels 1 min and 3 hr into the reperfusion period that the peak may have been missed. This is possible but seems unlikely. The rationale for the timepoints chosen was based on the report by Gurevitch et al. (1996) of a burst of TNF-a production immediately (1 min) upon myocardial reperfusion, whereas Squadrito et al. (1993) reported steadily increasing levels of TNF-a during the reperfusion period. Ultimately, methodological differences may explain the difference between these results and those of previous workers. For example, Gurevitch et al. (1996) and Squadrito et al. (1993) may have produced more severe myocardial damage than in our study because a more prolonged period of ischemia was used. In addition, these workers used a bioassay method to quantify TNF-a, whereas a highly sensitive ELISA method (using an antibody specific for rat TNF-a) was used in this study. Thus, although not definitive evidence, the measurements of serum TNF-a levels are hard to reconcile with the view that this cytokine plays an important role in lethal reperfusion injury in this rodent model. Consistent with this view, the selective PDE-IV inhibitor, rolipram, failed to reduce infarct size when administered 10 min prior to, and through the first 1 hr of reperfusion. In contrast, as a positive control, we demonstrated that ischemic preconditioning (Murry et al., 1986) was able to reduce infarct size by z41%. PDE-IV inhibition is known to be a highly effective strategy to inhibit TNF-a production by leukocytes (Molnar-Kimber et al., 1993; Prabhakar et al., 1994) and by myocardial tissue (Bergman and Holycross, 1996) and, consistent with this, rolipram is known to be a potent inhibitor of TNF-a production (Prabhakar et al., 1994; Semmler et al., 1993) and cardiac PDE-IV activity (Shahid and Nicholson, 1990). The dose of rolipram used was pharmacologically active and sufficient to inhibit PDE-IV and TNF-a production, because: (a) it was ten-fold higher than that previously shown to inhibit TNF-a production in rats (Prabahakar et al., 1994); and (b) it caused a small reduction in blood pressure. This reduction in blood pressure almost certainly reflects vasodilator activity, indicating that the bolus dose of rolipram achieved plasma levels of TNF-a high enough to cause some inhibition of PDE-III, which is present in vascular smooth muscle and mediates relaxation (Nicholson et al., 1991). Nevertheless, It is fair to point out that no direct evidence is presented to prove that the dose of rolipram used suppressed TNF-a production, although this was a difficult objective because of the unexpectedly low and declining levels of TNF-a that occurred in all groups during the reperfusion period. Finally, it is worth noting that the present data are highly consistent with the data of Simpson et al. (1992), who administered an almost identical dose of rolipram prior to reperfusion in a canine model of myocardial ischemia-reperfusion. These workers found that, although pharmacologically active plasma levels of rolipram were achieved, no reduction in infarct size was observed. Although this work does not find evidence in support of a role for TNF-a in myocardial reperfusion injury it is important to note two limitations of the present experiments. First, only circulating levels of TNF-a were measured. It is feasible that local levels of TNF-a in the myocardium were increased and contributed to the lethal cell injury observed, although the lack of effect of rolipram on infarct size is not consistent with this possibility. Further studies are required to address this question. Second, the present study used only necrosis as an endpoint for assessing reperfusion injury. Another form of reperfusion injury is myocardial stunning, a short-term func-

M. McVey et al. tional impairment of cardiac contractility. Interestingly, a recent report (Gurevitch et al., 1997) from studies using rat isolated perfused hearts suggests that locally produced TNF-a contributes to contractile dysfunction following ischemia-reperfusion. Thus, although the present results add to the literature regarding the role of TNF-a in lethal reperfusion injury, they do not clarify the importance of this cytokine in functional reperfusion injury (stunning). SUMMARY This study adds to the debate regarding the importance of TNF-a as a mediator of lethal reperfusion injury in the myocardium. The principal evidence presented is that measurements made with a specific ELISA demonstrate declining levels of TNF-a during the postischemia reperfusion period. In addition, a pharmacologically active dose of rolipram, an inhibitor of PDE-IV and TNF-a production, failed to reduce myocardial infarct size. Thus, collectively, the data from this study do not support an important pathophysiological role for TNF-a in myocardial necrosis in this rodent model of ischemiareperfusion injury. The authors thank Dr. Bruce Miller (RPR, Collegeville) for helpful advice regarding PDE-IV inhibitors and quantification of TNF-a levels in serum.

References Bergman M. R. and Holycross B. J. (1996) Pharmacological modulation of myocardial tumor necrosis factor a production by phosphodiesterase inhibitors. J. Pharmac. Exp. Ther. 279, 247–254. Chiu N., Park I. and Reid I. A. (1996) Stimulation of renin secretion by the phosphodiesterase IV inhibitor rolipram. J. Pharmac. Exp. Ther. 276, 1073–1077. Field G., Conn C. A., McClanahan T. B., Nao B. S., Kluger M. J. and Gallagher K. P. (1994) Tumor necrosis factor and interleukin-6 are not elevated in venous blood from ischemic canine myocardium. Proc. Soc. Exp. Biol. Med. 206, 384–391. Granger D. N. and Korthuis R. J. (1995) Physiologic mechanisms of postischemic tissue injury. Ann. Rev. Physiol. 57, 311–332. Gurevitch J., Frolkis I., Yuhas Y., Paz Y., Matsa M., Mohr R. and Yakirevich V. (1996) Tumor necrosis factor-alpha is released from the isolated heart undergoing ischemia and reperfusion. J. Am. Coll. Cardiol. 28, 247–252. Gurevitch J., Frolkis I., Yuhas Y., Lifschitz-Mercer B., Berger E., Paz Y., Matsa M., Kramer A. and Mohr R. (1997) Anti-tumor necrosis factor alpha improves myocardial recovery after ischemia and reperfusion. J. Am. Coll. Cardiol. 30, 1554–1561. Hansen P.R. (1995) Myocardial reperfusion injury: experimental evidence and clinical relevance. Eur. Heart J. 16, 734–740. Hearse D. J. and Bolli R. (1991) Reperfusion induced injury. Manifestations, mechanisms and clinical relevance. Trends Cardiovasc. Med. 1, 233–240. Herskowitz A., Choi S., Ansari A. A. and Wesselingh S. (1995) Cytokine mRNA expression in postischemic/reperfused myocardium. Am. J. Pathol. 146, 419–428. Jansen N. J. G., van Oeveren W., Gu Y. J., van Vliet M. H., Eijsman L. and Wildevuur C. R. H. (1992) Endotoxin release and tumor necrosis factor formation during cardiopulmonary bypass. Ann. Thorac. Surg. 54, 744–748. Maury C.P.J. and Teppo A.M. (1989) Circulating tumour necrosis factor-a (cachectin) in myocardial infarction. J. Int. Med. 225, 333-336. Miller T. D., Christian T. F., Hopfenspirger M. R., Hodges R. O., Gersh B. J. and Gibbons R. J. (1995) Infarct size after acute myocardial infarction measured by quantitative tomographic 99mTc Sestamibi imaging predicts subsequent mortality. Circulation 92, 334–341. Molnar-Kimber K., Yonno L., Heaslip R. and Weichman B. (1993) Modulation of TNF-a and IL-1b from endotoxin stimulated monocytes by selective PDE isozyme inhibitors. Agents Actions 39, C77–C79. Murry C. E., Jennings R. B. and Reimer K. A. (1986) Preconditioning with ischemia: a delay of lethal cell injury in ischemic myocardium. Circulation 74, 1124–1136. Nicholson C. D., Challiss R. A. J. and Shahid M. (1991) Differential modulation of tissue function and therapeutic potential of selective inhibitors of cyclic nucleotide phosphodiesterase isoenzymes. TIPS 12, 19–27. Pang X., Hershman J. M., Chung M. and Pekary A. E. (1989) Characterization of tumor necrosis factor-a receptors in human and rat thyroid cells

TNF-a and Myocardial Infarction and regulation of the receptors by thyotrophin. Endocrinology 125, 1783– 1788. Prabhakar U., Lipshutz D., O’Leary Bartus J., Slivjak M. J., Smith, E. F. III, Lee J. C. and Esser K. M. (1994) Characterization of cAMP-dependent inhibition of LPS-induced TNF-a production by rolipram, a specific phosphodiesterase IV (PDE IV) inhibitor. Int. J. Immunopharmac. 16, 805–816. Reeves M. L., Leigh B. K. and England P. J. (1987) The identification of a new cyclic nucleotide phosphodiesterase activity in human and guineapig cardiac ventricle. Biochem. J. 241, 535–541. Semmler J., Wachtel H. and Endres S. (1993) The specific type IV phosphodiesterase inhibitor rolipram suppresses tumor necrosis factor-a production by human mononuclear cells. Int. J. Immunopharmac. 15, 409–413. Shahid M. and Nicholson C. D. (1990) Comparison of cyclic nucleotide phosphodiesterase isoenzymes in rat and rabbit ventricular myocardium:

45 positive inotropic and phosphodiesterase inhibitory effects of Org 30029, milrinone and rolipram. Naunyn-Schmiedeberg’s Arch. Pharmac. 342: 698–705. Simpson P. J., Schelm J. A., Smallwood J. K., Clay M. P. and Lindstrom T.D. (1992) Inhibition of granulocyte cAMP-phosphodiesterase by rolipram in vivo is not sufficient to protect the canine myocardium from reperfusion injury. J Cardiovasc. Pharmac. 19, 987–995. Squadrito F., Altavilla D., Zingarelli B., Ioculano M., Calapai G., Campo G. M., Miceli A. and Caputi A. P. (1993) Tumor necrosis factor involvement in myocardial ischaemia-reperfusion injury. Eur. J. Pharmac. 237, 223–230. Van Bladel S., Libert C. and Fiers W. (1991) Interleukin-6 enhances the expression of tumor necrosis factor receptors on hepatoma cells and hepatocytes. Cytokine. 3, 149–154.