Hydrogen peroxide induces tumor necrosis factor α–mediated cardiac injury by a P38 mitogen-activated protein kinase–dependent mechanism

Hydrogen peroxide induces tumor necrosis factor α–mediated cardiac injury by a P38 mitogen-activated protein kinase–dependent mechanism Daniel R. Meldrum, MD, Charles A. Dinarello, MD, Joseph C. Cleveland, Jr., MD, Brian S. Cain, MD, Brian D. Shames, MD, Xianzhong Meng, MD, PhD, and Alden H. Harken, MD, Denver, Colo.

Background. Oxidant stress caused by ischemia or endotoxemia induces myocardial dysfunction and cardiomyocyte death; however, mechanisms responsible remain unknown. We hypothesized that hydrogen peroxide (H2O2) induces myocardial dysfunction and cardiomyocyte death via P38 mitogen-activated protein kinase (MAPK)–mediated myocardial tumor necrosis factor (TNF) production. Methods. Langendorff perfused rat hearts (6/group) were subjected to oxidant stress (H2O2 infusion; 300 mmol/L × 80 minutes), with and without prior infusion of a specific P38 kinase MAPK inhibitor (P38i = 1 mmol/L/min × 5 minutes) or TNF neutralization (20 mg TNF binding protein (BP)/min × 80 minutes). Developed pressure (DP), coronary flow, and end-diastolic pressure were continuously recorded. Myocardial creatine kinase (CK) loss was measured in the coronary effluent, and tissue TNF was measured in myocardial homogenates. Results. Eighty minutes of H2O2 infusion induced a 6.5-fold increase in myocardial TNF production, which was associated with a 70% decrease in DP and increase in CK loss. P38 MAPK inhibition or TNF-BP decreased myocardial TNF production, cardiomyocyte death, and myocardial dysfunction. Conclusions. These results demonstrate that H2O2 alone induces myocardial TNF production. P38 MAPK is an oxidant-sensitive enzyme that mediates oxidant-induced myocardial TNF production, cardiac dysfunction, and cardiomyocyte death. (Surgery 1998;124:291-7.) From the Departments of Surgery and Medicine, Divisions of Cardiothoracic Surgery and Infectious Diseases, University of Colorado Health Sciences Center, Denver, Colo.

EVIDENCE INDICATES THAT cytokines are important mediators of cardiovascular disease.1-3 Tumor necrosis factor-α (TNF) is a proinflammatory cytokine that has been implicated in the pathogenesis of septic-associated, traumatic-associated, and hypovolemic shock-associated cardiac dysfunction, as well as cardiovascular diseases including acute myocardial infarction, chronic heart failure, atherosclerosis, viral myocarditis, and cardiac allograft rejection.4 The heart contains resident macrophages5 and is a rich source of TNF.6

Supported by National Institutes of Health Grants HL-43696, HL-44186, GM-08315 (A.H.H.), AI-15614 (C.A.D.) and the National Institutes of Health National Research Service Award (D.R.M.). Presented at the Fifty-ninth Annual Meeting of the Society of University Surgeons, Milwaukee, Wis., Feb. 12-14, 1998. Reprint requests: Daniel R. Meldrum, MD, Division of Cardiothoracic Surgery, University of Colorado Health Sciences Center, 4200 East Ninth Avenue, Box C-306, Denver, CO 80262. Copyright © 1998 by Mosby, Inc. 0039-6060/98/$5.00 + 0 11/6/90570

Cardiac myocytes themselves also produce TNF.1-3,5,7 In fact, in response to endotoxin, the myocardium produces as much TNF per gram tissue as either the liver or the spleen, both possessing large numbers of macrophages, and are major sources of TNF.5 Kapadia et al5 demonstrated that myocardial TNF production is evenly distributed between cardiomyocytes and cardiac macrophages. Thus, local myocardial TNF appears to be an important source of TNF affecting myocardial function. We have recently demonstrated that ischemia and reperfusion injury induces myocardial TNF production in an isolated rat heart preparation3; the mechanisms responsible remain unknown. Reperfusion of ischemic myocardium imposes an oxidant burden that directly injures myocardium8-10; however, oxidation products (hydrogen peroxide) may also activate oxidantsensitive enzymes (eg, P38 mitogen-activated protein kinase [MAPK]) involved in initiating cardiac TNF production.11,12 Thus we hypothesized that oxidant stress alone is sufficient to SURGERY 291

292 Meldrum et al.

induce myocardial TNF production by a P38 MAPK-dependent mechanism. MATERIAL AND METHODS Material. Male Sprague-Dawley rats (weight, 325 to 350 g; Sasco Inc., Omaha, Neb.) were fed a standard diet and acclimated for 2 weeks before the experiments. The animal protocol was approved by the Animal Care and Research Committee of the University of Colorado Health Sciences Center. Animals received care in compliance with the “Guide for the Care and Use of Laboratory Animals” (NIH publication no. 85-23, revised 1985). Unless otherwise stated, reagents were obtained from Sigma Chemical Co, St Louis, Mo. Perfusion of the isolated rat heart: developed pressure, end-diastolic pressure, coronary flow, and heart rate measurements. The isolated crystalloid-perfused rat heart model was used as previously described.13 In brief, after anesthesia and heparinization (sodium pentobarbital, 60 mg/kg intraperitoneally and heparin-sodium, 500 units intraperitoneally) hearts were excised into 4°C Krebs-Henseleit and perfused with oxygenated buffer within 45 seconds. Hearts then were retrograde perfused in the isolated, isovolumetric Langendorff mode (70 mm Hg) with modified Krebs-Henseleit solution (in mmol/L: 5.5 glucose, 1.2 CaCl2, 4.7 KCl, 25.0 NaHCO3) and saturated with 92.5% O2/7.5% CO2, to achieve a PO2 of 440 to 460 mmHg, PCO2 of 39 to 41 mmHg, and pH of 7.39 to 7.41 (ABL-4 blood gas analyzer; Radiometer, Copenhagen, Denmark). A pulmonary arteriotomy and left atrial resection were performed before insertion of a water-filled latex balloon through the left atrium into the left ventricle. The balloon was then adjusted to a left ventricular end-diastolic pressure (EDP) of 6 mm Hg during the initial equilibration. The preload volume was held constant during the entire experiment to allow continuous recording of the ventricular pressure. Pacing wires were fixed to the right atrium and pulmonary outflow tract, and hearts were paced at approximately 6 Hz (355 beats per minute) to compare functional measurements using a standardized heart rate. The measured indexes of myocardial function were left ventricular developed pressure (DP), EDP and coronary flow (CF). Data were continuously recorded using a MacLab 8 preamplifier/digitizer (AD Instruments Inc., Milford, Mass) and an Apple Quadra 800 computer (Apple Computer Inc., Cupertino, Calif). Developed pressure is a wellaccepted index of the strength of myocardial contraction, which, at fixed rates, correlates with

Surgery August 1998

cardiac output in the working heart preparation. A 3-way stopcock above the aortic root was for continuous H2O2 or drug infusions. Coronary flow was measured by collecting pulmonary artery effluent. After perfusion, left ventricular myocardium was excised and added to 10 volumes of cold isotonic homogenization buffer (in mmol/L: 50 imidazole acetate, 10 mg acetate, 4 KH2PO4, 2 ethylenediaminetetraacetic acid; and in µmol/L: 50 N-acetylcysteine as antioxidant and 12.5 sulfur in 0.8% EtOH to inhibit adenylate cyclase; pH 7.6). Samples were homogenized in a vertishear tissue homogenizer (parallel blades 0.5 cm apart) at half maximal speed for 20 s (10 equally spaced bursts) followed by centrifugation at 2000g for 15 minutes (4°C) and then ultrasonicated. Supernatant total protein concentration was measured using the Lowry assay against a bovine serum albumin standard and then stored at –70°C until used in the TNF assays. Experimental design and groups. Hearts that did not maintain an equilibration DP of 85 mm Hg were discarded. Injury control hearts underwent 80 minutes of normothermic perfusion with standard perfusate containing 300 mmol/L of hydrogen peroxide (H2O2). The P38 MAPK inhibitor (SB 203580; gift from Dr. John Lee, SmithKline Beecham) was infused at a concentration of 1 mmol/L × 5 minutes before the H2O2 infusion. SB 203580 is a pyridinyl imidazole. Pyridinyl imadazoles, which abolish proinflammatory monokine production, are highly selective inhibitors of P38 MAPK and its downstream products, but not of Ras or Raf-1, activation.14 The dose chosen is based on previous work by Shapiro and Dinarello,15 which demonstrated that this dose effectively inhibits monocyte cytokine production. Further, the dose is based on dose response curves generated in this laboratory, which demonstrated that the dose of SB 203580 effectively inhibits ischemia-induced TNF production.16 In a separate group, TNF-binding protein (TNF-BP; 20 mg/min × 80 minutes) was administered throughout the H2O2 infusion period. SB 203580 was dissolved in dimethyl sulfoxide and diluted in normal saline containing 0.25% (w/v) human serum albumin (Abbott Laboratories, North Chicago, Ill) vehicle. Recombinant human TNF-BP was supplied by Dr Carl Edwards, Amgen, Inc, Boulder, Colo. TNF-BP is expressed in Escherichia coli as the four extracellular domains of the p55 TNF receptor.17 TNF-BP was diluted in normal saline containing 0.25% human serum albumin. Control hearts received

Surgery Volume 124, Number 2

either vehicle pretreatment with and without subsequent H2O2 infusion. Myocardial TNF. Myocardial homogenate TNF content was determined by enzyme-linked immunosorbent assay (ELISA; Genzyme, Cambridge, Mass). The ELISA was performed by adding 100 ml of each sample (equal protein and tested in duplicate) to wells in a 96-well plate of a commercially available ELISA kit. According to the manufacturer, mouse anti-TNF antibodies used in this ELISA are not influenced by either the type 1 or type 2 TNF receptors (TNFR1 or TNFR2, respectively) and have a lower limit of detectability 50 pg/mL. TNF ELISA was performed according to the manufacturer’s instructions. Final results were expressed as picograms of TNF per gram. Coronary effluent creatine kinase activity. Coronary effluent (1 mL) was collected during equilibration and at 10, 20, 30, and 40 minutes into reperfusion and then frozen at –70°C until assay. All assays were performed within 2 weeks of effluent collection. The assay was performed with Sigma diagnostic kit no. 47-UV on an automated spectrophotometer (Centrifichem 500 discrete auto-analyzer; Union Carbide) in cuvettes maintained at 30°C. Samples and reagents were maintained at 4°C before assay. Solutions were prepared in distilled deionized water. Results are presented as units/L CK activity. Presentation of data and statistical analysis. All reported values are mean ± SEM (n = 6/group). Differences at the 95% confidence level were considered significant. Data were compared at the corresponding time points between groups using one way analysis of variance (ANOVA) with post hoc Bonferroni/Dunn test (StatView 4.0; Abacus Concepts, Berkeley, Calif). RESULTS Myocardial TNF. H2O2 infusion resulted in 6.5fold increase in myocardial TNF concentration (Fig. 1). Pretreatment with the P38 MAPK inhibitor SB 203580 resulted in a reduction in oxidantinduced TNF production (from 1.3 ± 0.2 to 0.4 ± 0.11 ng/g). TNF neutralization (TNF-BP) decreased oxidant-induced TNF production (0.7 ± 0.17 ng/g) to a lesser extent. Myocardial function. Eighty minutes of H2O2 infusion resulted in a 70% decrease (P < .05) in myocardial function; DP decreased from 104 ± 7 to 33 ± 6 mm Hg (Fig. 2); coronary flow (CF) decreased from 19 ± 1.7 to 13 ± 2 mL/min (Fig. 3); and EDP increased (decreased compliance) from 4 ± 2.9 to 57 ± 8 mm Hg (Fig. 4). Inhibition of myocardial TNF production by SB 203580 or with TNF-BP improved

Meldrum et al. 293

Fig. 1. Myocardial TNF after H2O2-induced myocardial injury. Eighty minutes of H2O2 (300 mmol/L) infusion, under otherwise normothermic perfusion conditions, induced myocardial TNF production. Control hearts (CTRL) were subjected to 80 minutes of normothermic perfusion without H2O2 in the perfusate. P38 MAPK inhibition (H2O2+P38i) or TNF neutralization (H2O2+TNFBP) reduced H2O2-induced myocardial TNF production. *P < .05 vs control; **P < .05 vs H2O2; n = 6/group.

Fig. 2. Left ventricular DP after H2O2-induced myocardial injury. Eighty minutes of H2O2 (300 mmol/L) infusion induced myocardial contractile dysfunction. Control hearts (CTRL) were subjected to 80 minutes of normothermic perfusion without H2O2 in the perfusate. P38 MAPK inhibition (H2O2+P38i) or TNF neutralization (H2O2+TNF-BP) reduced H2O2-induced myocardial contractile dysfunction. *P < .05 vs control; **P < .05 vs H2O2; n = 6/group.

myocardial function as follows: DP increased to 67 ± 4 and 56 ± 6 mm Hg, respectively; CF increased to 17 ± 1.5 and 16 ± 1.2 mL/min, respectively; and EDP decreased to 32 ± 3 and 37 ± 4 mm Hg, respectively. Myocardial creatine kinase leak. Creatine kinase (CK) activity in the coronary effluent was undetectable during equilibration but increased to 84 ± 11 units/mL after 80 minutes of H2O2 infusion (Fig. 5). Inhibition of myocardial TNF production by SB 203580 or with TNF-BP attenuated oxidant-induced myocardial CK loss to 39 ± 8 and 51 ± 9 units/mL, respectively.

294 Meldrum et al.

Fig. 3. CF after H2O2-induced myocardial injury. Eighty minutes of H2O2 (300 mmol/L) infusion reduced CF despite constant perfusion pressure. Control hearts (CTRL) were subjected to 80 minutes of normothermic perfusion without H2O2 in the perfusate. P38 MAPK inhibition (H2O2+P38i) or TNF neutralization (H2O2+TNFBP) improved CF after H2O2 infusion. *P < .05 vs control; **P < .05 vs H2O2; n = 6/group.

Fig. 4. Left ventricular EDP during H2O2-induced myocardial injury. Eighty minutes of H2O2 (300 mmol/L) infusion induced diastolic dysfunction. Control hearts (CTRL) were subjected to 80 minutes of normothermic perfusion without H2O2 in the perfusate. P38 MAPK inhibition (H2O2+P38i) or TNF neutralization (H2O2+TNF-BP) improved EDP after H2O2 infusion. *P < .05 vs control; **P < .05 vs H2O2; n = 6/group.

DISCUSSION These results demonstrate that H 2O 2 itself increases myocardial TNF production and cardiac dysfunction via a P38 MAPK-dependent mechanism. Oxidant stress provoked myocardial TNF production, even in the absence of global ischemia and reperfusion. These observations are consistent with the concept that cardiac myocytes and cardiac resident macrophages are important sources of TNF that affect myocardial function. These findings also indicate that oxidant-induced myocardial TNF adversely affects myocardial function because inhibition of TNF production (P38

Surgery August 1998

MAPK inhibitor, SB 203580) or TNF neutralization (TNF-BP) reduced oxidant-induced myocardial damage. Previously we have demonstrated that ischemia and reperfusion induce TNF production in this model. 3 These results may suggest that the oxidant component of ischemia and reperfusion injury plays an important role in myocardial TNF production after such conditions. This may have important clinical implications for preservation of myocardium during and after ischemia. Heart transplant surgery and cardiac bypass surgery, as well as coronary angioplasty, are three clinical scenarios that obligate myocardial ischemia-reperfusion. TNF has recently been appreciated as an important mediator of myocardial ischemic damage.6,18 The present model may allow the study of these mechanisms in the laboratory, as well as the opportunity to test potential therapeutic strategies. Raf-1/MEK appears to activate members of the MAPK family of protein kinases. Of these the P38 MAPK appears to be a pivotal MAPK in the cascade leading to TNF gene induction.15,19,20 Han et al21 isolated a 38 kd macrophage protein kinase that was phosphorylated after lipopolysaccharide (LPS) treatment. Subsequent cloning and sequence analysis19,22 revealed that the novel MAPK homologue was closely related to the osmolar sensitive Hog-1 gene in yeast. During the same period, Lee et al20 were searching for the target proteins of a novel class of anti-inflammatory drugs (pyridinyl imidazoles) capable of abolishing LPS-induced proinflammatory monokine production. Photoaffinity labeling of drug analogues identified proteins that proved to be the mammalian equivalent of the Hog-1 gene product. The 38 kd protein was identical to the MAPK cloned by Han et al.19,22 Pyridinyl imadazoles, which abolish proinflammatory monokine production, are highly selective inhibitors of P38 MAPK and its downstream products, but not of Ras or Raf-1, activation.14 Huot et al11 and Guyton et al12 have independently demonstrated that H2O2 activates P38 MAPK. This study sought to extend these observations and to relate them to myocardial function. It is well established that TNF induces potent cardiovascular effects. For example, TNF induces hemodynamic alterations including decreased ejection fraction and reduced myocardial contractile efficiency, biventricular dilation, and decreased systemic vascular resistance and hypotension. TNF also depresses myocardial function in an ex vivo, crystalloid-superfused papillary muscle preparation.23 Although TNF is believed to mediate LPSinduced myocardial depression, ischemia-

Surgery Volume 124, Number 2

provoked myocardial TNF production has only recently been reported.3 Ischemia-induced myocardial TNF production is likely more common clinically than sepsis-induced myocardial TNF by several orders of magnitude. The mechanisms by which TNF causes myocardial dysfunction include calcium dyshomeostasis, direct cytotoxicity, oxidant stress, disruption of excitation-contraction coupling, myocyte apoptosis, and induction of other cardiac depressants. The biphasic nature of TNF-induced myocardial depression suggests that TNF induces negative inotropic effects by at least two different mechanisms.1,4 The early phase of TNF-induced functional depression occurs within minutes, whereas the delayed phase appears to require hours of TNF exposure.4 Although nitric oxide (NO) production has been demonstrated to mediate TNF-induced depression in some models, TNF may not induce high levels of NO rapidly enough to account for the early phase of myocardial depression.4 Sphingolipid metabolites are stress-induced second messengers that participate in intracellular signal transduction after TNF binding to the TNF receptor type 1 (TNFR1; 55 kd). Two characteristics of sphingolipid metabolites suggested1,4 that sphingosine mediates TNF-induced myocardial contractile dysfunction: (1) it is rapidly produced by cardiac myocytes (via sphingomyelin degeneration) after TNF triggering of TNFR1, and (2) sphingosine decreases calcium transients by blocking the ryanodine receptor, which mediates calcium-induced calcium release from the sarcoplasmic reticulum.24 Oral et al4 reported that myocardial sphingosine production occurred rapidly after TNF administration and temporally correlated with dysfunction and calcium dyshomeostasis of cardiac myocytes. Inhibition of sphingosine production abolished TNF induced contractile dysfunction, and sphingosine administration mimicked TNF-induced contractile depression. TNF also may induce contractile dysfunction by inducing apoptosis.18 Krown et al25 demonstrated that TNF induced cardiocyte apoptosis by a sphingosine-dependent mechanism. Considerable information exists concerning the mechanisms by which LPS induces TNF production; however, little is known about the mechanisms of ischemia-induced myocardial TNF production. Reperfusion of ischemic myocardium likely imposes an oxidant burden in which the reduction product of molecular oxygen, H2O2, contributes to myocardial injury.8 H2O2 has been demonstrated to activate P38 MAPK in other models, which may contribute to ischemia-induced

Meldrum et al. 295

Fig. 5. Myocardial CK loss during H2O2-induced myocardial injury. Eighty minutes of H2O2 (300 mmol/L) infusion induced myocardial CK loss. Control hearts (CTRL) were subjected to 80 minutes of normothermic perfusion without H2O2 in the perfusate. P38 MAPK inhibition (H2O2+P38i) or TNF neutralization (H2O2+TNF-BP) reduced H2O2-induced myocardial CK loss. *P < .05 vs control; **P < .05 vs H2O2; n = 6/group.

TNF production.11,12 Indeed, the results of this study suggest that H2O2 alone induces myocardial TNF production by a P38 MAPK-dependent mechanism. It must be recognized, however, that other oxidant-sensitive enzymes may play a role in this process. We have previously demonstrated that hemorrhage and resuscitation, albeit a complex stimulus, also activate oxidant-sensitive transcription factor nuclear factor kappa B (NF-κB), an important transcription factor for TNF production. It remains to be determined whether H2O2 activates myocardial NF-κB, which may play a role in the sequence of signaling events leading to oxidant-induced myocardial TNF production. TNF is a proinflammatory cytokine. Proinflammatory cytokines act to increase their own production and the synthesis of small inflammatory mediators such as platelet-activating factor, eicosanoids, and oxidative radicals. Proinflammatory cytokines also recruit and stimulate cellular components of the immune system. Because TNF is a proinflammatory cytokine, it is a potent stimulant of it own production (ie, it works to enhance its own production in a positive feedback fashion). Disruption of this positive feedback cycle with TNFBP dramatically decreases not only TNF activity, but also its production. Indeed, results of this study indicate that TNF-BP decreases myocardial TNF production. However, it is also possible that the decreased TNF is simply a reflection of TNF sequestration by the binding protein. Several lines of evidence suggest that TNF participates in myocardial ischemia-reperfusion injury.1,4 Thus, strategies

296 Meldrum et al.

designed to decrease myocardial TNF production may be of therapeutic value. Although this model may provide some clues regarding the mechanisms of oxidant-induced myocardial dysfunction, the study should be interpreted with several important caveats. The ex vivo myocardial perfusion model used here does not precisely recreate the in vivo situation. We have successfully used this model to answer questions concerning myocardial physiology and pathophysiology in the past; however, in many ways it does not reflect the in vivo situation. Because this heart model is perfused with crystalloid, not blood, the formed elements of plasmaderived protective antioxidants or injurious platelets and phagocytes do not contribute. Indeed, this model was chosen to more accurately determine potential mechanisms of myocardial TNF production (vs production from blood cells, and in the absence of the complex stimulus of global ischemia and reperfusion). Furthermore, oxidant-induced myocardial TNF production almost certainly does not account for the entire injury observed in this model. Although TNF blockade provided significant improvement, it did not completely abolish injury. Other mediators and direct mechanisms of injury are likely involved. We thank Dr John Lee (SmithKline and Beecham) for providing some of the SB 203580 for the preliminary experiments and Dr Carl Edwards (Amgen, Inc) for providing the TNF-BP. REFERENCES 1. Meldrum DR. Tumor necrosis factor in the heart (review). Am J Physiol 1998;274:R577-95.. 2. Meldrum DR, Shenkar R, Sheridan BC, Cain BS, Abraham E, Harken AH. Hemorrhage activates myocardial NFkB and increases tumor necrosis factor in the heart. J Mol Cell Cardiol 1997;29:2849-54. 3. Meldrum DR, Cain BS, Cleveland JC, Meng X, Ayala A, Banerjee A, et al. Adenosine decreases post-ischemic myocardial TNFα production: anti-inflammatory implications for preconditioning and transplantation. Immunology 1997;92:472-7. 4. Oral H, Dorn GW, Mann DL. Sphingosine mediates the immediate negative inotropic effects of tumor necrosis factora in the adult mammalian cardiac myocyte. J Biol Chem 1997;272:4836-42. 5. Kapadia S, Lee J, Torre-Amione G, Birdsall HH, Ma TS, Mann DL. Tumor necrosis factor-a gene and protein expression in adult feline myocardium after endotoxin administration. J Clin Invest 1995;96:1042-52. 6. Wan S, DeSmet JM, Barvais L, Golstein M, Vincent JL, LeClerc JL. Myocardium is a major source of proinflammatory cytokines in patients undergoing cardiopulmonary bypass. J Thorac Cardiovasc Surg 1996;112:806-11. 7. Meng X, Ao L, Brown JM, Meldrum DR, Sheridan BC, Cain BS, et al. LPS induces delayed cardiac functional protection against ischemia independent of cardiac and circulating TNFα. Am J Physiol In press.

Surgery August 1998 8. Brown JM, Terada LS, Grosso MA, Whitman GJ, Velasco SE, Patt A, et al. Xanthine oxidase produces H2O2 which contributes to reperfusion injury of ischemic isolated rat hearts. J Clin Invest 1988;81:1297-301. 9. Meldrum DR, Cleveland JC, Rowland RT, Banerjee A, Harken AH, Meng X. Early and delayed preconditioning: differential mechanisms and additive protection. Am J Physiol 1997;273:H725-33. 10. Meldrum DR, Cleveland JC, Mitchell MB, Sheridan BC, Robertson F, Harken AH, et al. Protein kinase C mediates Ca2+ induced cardioadaptation to ischemia-reperfusion injury. Am J Physiol 1996;271:R1718-26. 11. Huot J, Houle F, Marceau F, Landry J. Oxidative stressinduced actin reorganization mediated by the p38 mitogen-activated protein kinase/heat shock pathway in vascular endothelial cells. Circ Res 1997;80:383-92. 12. Guyton KZ, Liu Y, Gorospe M, Xu Q, Holbrook NJ. Activation of mitogen-activated protein kinase by H2O2: role in cell survival following oxidant injury. J Biol Chem 1996;271:4138-42. 13. Meldrum DR, Cleveland JC, Sheridan BC, Rowland RT, Banerjee A, Harken AH. Cardiac preconditioning with calcium: clinically accessible myocardial protection. J Thorac Cardiovasc Surg 1996;112:778-86. 14. Lee JC, Young PR. Role of CSB/p38/RK stress response kinase in LPS and cytokine signaling mechanisms. J Leukoc Biol 1996;59:152-7. 15. Shapiro L, Dinarello CA. Osmotic regulation of cytokine synthesis in vitro. Proc Natl Acad Sci 1995;92:12230-4. 16. Meldrum DR, Dinarello CA, Meng X, Shapiro L, Harken AH. P38 MAP kinase-mediated myocardial TNFα production contributes to post-ischemic cardiac dysfunction. Circulation 1997;96:I-556. 17. Engelman H, Aderka D, Rubinstein M, Rotman D, Wallach D. A tumor necrosis factor-binding protein purified to homogeneity from human urine protects cells from tumor necrosis factor toxicity. J Biol Chem 1989;264:11974-80. 18. Olivetti G, Abbi R, Quaini F, Kajstura J, Cheng W, Nitahara JA, et al. Apoptosis in the failing human heart. N Engl J Med 1997;336:1131-41. 19. Han J, Richter B, Li Z, Kravchenko V, Ulevitch RJ. Molecular cloning of the p38 MAP kinase. Biochim Biophys Acta 1995;16:224-7. 20. Lee JC, Laydon JT, McDonnel PC, Gallagher TF, Kumar S, Green D, et al. A protein kinase involved in the regulation of inflammatory cytokine biosynthesis. Nature 1994;372:739-46. 21. Han H, Lee JD, Tobias PS, Ulevitch RJ. Endotoxin induces rapid tyrosine phosphorylation in 70Z/3 cells expressing CD14. J Biol Chem 1993;268:25009-14. 22. Han J, Lee JD, Bibbs L, Ulevitch RJ. A MAP kinase targeted by endotoxin and hyperosmolarity in mammalian cells. Science 1994;265:808-11. 23. Finkel MS, Oddis CV, Jacob TD, Watkins SC, Hattler BG, Simmons RL. Negative inotropic effects of cytokines on the heart mediated by nitric oxide. Science 1992;257: 387-9. 24. Sabbadini RA, Betto R, Teresi A, Fachechi-Cassano G, Salviati G. The effects of sphingosine on sarcoplasmic reticulum membrane calcium release. J Biol Chem 1992;267:15475-84. 25. Krown KA, Page MT, Nguyen C, Zechner D, Gutierrez V, Comstock KL, et al. Tumor necrosis factor alpha-induced apoptosis in cardiac myocytes: involvement of the sphingolipid signaling cascade in cardiac cell death. J Clin Invest 1996;98:2854-65.

Surgery Volume 124, Number 2 DISCUSSION Dr Joseph B. Zwischenberger (Galveston, Texas). You are trying to study the effects of an oxidant stress alone. Yet, in your model on the oxidant stress group, EDP increases from 4 mm Hg, which is normal, to 57 ± 8 mm Hg. When you give the P38 MAPK, it increases to 32 mm Hg. With such an outrageous pressure on the EDP, with both pressure and stretch, you may in fact be looking at an ischemic or a stretch injury to the myocardium alone. Dr Meldrum. The induction of the injury by oxidant stress results purely from the infusion of the H2O2, and the initiation of the injury is purely oxidant stress alone. Once the injury has been initiated, I agree that decreased flow is probably a factor in potentiating further injury. Dr Timothy R. Billiar (Pittsburgh, Pa.). I think this is an important study because it has implications not just to myocardial ischemia/reperfusion but to ischemia/reperfusion in essentially any situation (eg, stroke, shock). The important message is that oxidant-induced injury is not necessarily just a direct toxicity of oxygen radicals and their products to the tissues. It also results from activation of specific inflammatory pathways, and it is this redox activation of inflammation that is so important. Your group has worked with ischemia/reperfusion in the past. Have you been able to demonstrate a specific P38/NF-κB/TNF

Meldrum et al. 297

pathway in an ischemia/reperfusion model? My second question has to do with the mechanism of activation. You have emphasized that P38 MAPK is the redox-sensitive protein, but I am not sure that the literature supports this assumption. Perhaps the redox sensors may be upstream of P38. Ras would be an example. What do you think these redox switches might be? Dr Meldrum. We chose this model to determine what effect oxidant stress alone may have on the heart in the absence of blood or other confounding variables. We also wanted to evaluate whether the heart is capable of producing TNF by itself, in the absence of circulating blood stimulation or other organ effects. In this isolated rat heart preparation, we saw an increase in myocardial TNF. We have also looked at myocardial TNF after hemorrhage or resuscitation. We have recently observed that hemorrhage and resuscitation do induce NF-κB activation within the heart as demonstrated by a mobility shift assay. We are currently looking at ischemia/reperfusion-induced NF-κB activation. With regard to other oxidant sensors, we looked at P38 MAPK, but many others, including Ras, probably are important upstream oxidant sensors, as well as NF-κB downstream, P38 MAPK, and inducible NO synthase. Those are all important, as they may offer potential therapeutic targets.