Volatile anesthetics are routinely used for general anesthesia. Experimentally they protect the myocardium against ischemia/reperfusion (IR) injury when administered either before or after prolonged ischemia (1,2). The unique discovery that short periods of ischemia before prolonged ischemia precondition the myocardium to attenuate IR injury, called ischemic preconditioning (IPC) (3), was followed by studies showing that temporary exposure to volatile anesthetics before prolonged ischemia also preconditioned against IR injury. This phenomenon, called “anesthetic preconditioning” (APC), reduced infarct size in dogs (4) and improved vascular, mechanical, and metabolic function, reduced cytosolic Ca2+ overload and nicotinamide-adenine dinucleotide (NADH) accumulation, and improved drug-induced endothelial nitric oxide release in guinea pig isolated hearts (5–7).
The precise triggering mechanism of APC is not well understood. Attenuation of APC by the adenosine triphosphate (ATP)-sensitive K+ (KATP) channel antagonist glibenclamide (4,5) suggested that volatile anesthetics trigger APC, in part, by opening KATP channels. Evidence that KATP channels in the mitochondrial (m) inner membrane might also play an important role in preconditioning was first presented by Garlid et al. (8).
Mitochondria transport Ca2+ to regulate m[Ca2+], which controls the activation of Ca2+-sensitive dehydrogenases and oxidative phosphorylation (9). Overload of mCa2+ occurs during ischemia and on reperfusion and has been causally implicated in IR injury (10–12). IPC reduces mCa2+ overload during IR, an effect that was blocked by the putative mKATP channel antagonist 5-hydroxydecanoate (5-HD) (13). Mitochondrial KATP channel opening is thought to reduce the driving force for intracellular Ca2+ into the mitochondria (11,14). The possible influence of altered mCa2+ in APC and the effect of opening mKATP channels on mCa2+ have not been examined.
We hypothesized that anesthetic exposure before ischemia decreases subsequent mCa2+ overload and attenuates IR injury in part by opening mKATP channels. Use of the Langendorff heart preparation with real-time measurement of mCa2+ fluorescence (F) allowed us to follow the development of functional variables, as well as mCa2+, before, during, and after ischemia in guinea pig intact hearts. We examined whether APC is evidenced by reduced mCa2+ during IR and whether 5-HD reverses the cardioprotection afforded by APC.
The investigation conformed to the Guide for the Care and Use of Laboratory Animals (US National Institutes of Health No. 85-23, revised 1996) and was approved by the animal studies committee. Our methods have been described previously (2,5–7,12). In brief, 40 hearts were isolated from decapitated guinea pigs and perfused retrograde through the aorta at a constant pressure of 55 mm Hg and at 37°C. The Krebs-Ringer solution (KR) perfusate (pH 7.40 ± 0.01; Po2, 570 ± 10 mm Hg) had the following composition (in mM, non-ionized): Na+ 138, K+ 4.5, Mg2+ 1.2, Ca2+ 2.5, CI− 134, HCO3 14.5, H2PO4 1.2, glucose 11.5, pyruvate 2, mannitol 16, probenecid 0.1, EDTA 0.05, and insulin 5 (U/L).
Heart rate, left ventricular pressure (LVP), and coronary inflow (ultrasonic flowmeter) were measured as described previously (2,5–7,12). Characteristic data from LVP were systolic (sys), diastolic, and developed LVP, as well as dLVP/dtmax and dLVP/dtmin (indices of contractility and relaxation, respectively). Coronary inflow and coronary venous Na+, K+, Ca2+, Po2, pH, and Pco2 were measured off-line. Sevoflurane was bubbled into the perfusate for 15 min by using an agent-specific vaporizer placed in the oxygen/CO2 gas mixture line. Coronary perfusate was collected from a port in the aortic cannula and from the pulmonary catheter to measure sevoflurane concentrations by gas chromatography. The inflow sevoflurane concentration was 1.24 ± 0.02 mM, or 8.80 ± 0.14 vol%, at 37°C (n = 20). This concentration, which is too large for maintenance of clinical anesthesia but can be used temporarily during mask induction (15), was chosen to unmask any possible effect of sevoflurane on mCa2+ during exposure to sevoflurane as well as during IR. To block mKATP channels (8), 200 μM 5-HD was infused alone or before, during, and after sevoflurane exposure.
At the end of 120 min of reperfusion, hearts were removed and ventricles cut into transverse sections of 3-mm thickness. Sections were immediately stained with 1% 2,3,5-triphenyletrazolium chloride in 0.1 M KH2PO4 buffer (pH 7.4, 38°C) for 5 min (7). All slices were digitally imaged, and the infarcted areas were measured in a blinded fashion by planimetry with National Institutes of Health (Bethesda, MD) software (Image 1.62). Infarcted areas of individual slices were averaged on the basis of their weight to calculate the total infarct size of both ventricles as a percentage.
The m[Ca2+] was measured by the indo-1 F technique (16) adapted for use in the intact heart (12,17,18). Each experiment was performed in a light-blocking Faraday cage. The distal end of a bifurcated fiberoptic cable (6.8 mm2 per bundle) was placed gently against the left ventricular anterior wall. A net was applied around the heart to ensure optimal contact with the fiberoptic tip. The two proximal ends of the fiberoptic cable were connected to a modified luminescence spectrophotometer (Photon Technology International [PTI], London, Ontario, Canada).
F was excited with light from a 75-W xenon arc lamp. The light was filtered at 350 nm (Delta RAM; PTI), and the beam was focused onto the fibers entering the optic bundle. The arc lamp shutter was opened for only 2.5-s recording intervals to prevent photobleaching. F collected by the second limb of the cable was separated by a dichroic beamsplitter at 430 nm and filtered at 405 ± 15 nm and 460 ± 10 nm. Intensities were measured by photomultipliers (Photomultiplier Detection System 814; PTI). After a 30-min equilibration period, background F was determined for each heart. Indo-1-AM (6 μM) was prepared (2,6,12) and perfused for 30 min. This loading increased F approximately 10-fold.
After washout of residual interstitial indo-1-AM for 20 min, cytosolic Ca2+ was quenched from cytosolic indo-1 by perfusion with 100 μM MnCl2 for 15 min. The remaining F originates predominantly from mitochondrial sources (18) and lacks the phasic character of cytosolic Ca2+ F (12,13). Figure 1 displays a representative original F tracing before and after indo-1 loading, after quenching with MnCl2, and 200 min later in a nonischemic control heart. Although both F405 and F460 declined over time, the F405/F460 ratio remained stable during the course of our studies (6,12).
Background F405 and F460 values obtained before indo-1 loading represent the autofluorescence of mNADH (7,12,19). The m[Ca2+] was corrected for the corresponding drug- and IR-induced changes in autofluorescence (NADH) obtained previously (7). The indo-1 transient is nonlinearly proportional to m[Ca2+], which was calculated according to the following equation (12,17,18):MATH where S460 is the ratio of F intensities at 460 nm at zero and saturated Ca2+, Kd is the dissociation constant of indo-1, Rm is the actually measured F405/F460 ratio, Rmin is the F405/F460 ratio at zero Ca2+, and Rmax is the F405/F460 ratio at saturated Ca2+. In calibration experiments, Kd was calculated as 249 nM (12), S460 as 2.29, Rmin as 0.57, and Rmax as 6.22 at the chosen photomultiplier settings; m[Ca2+] is given in nanomolar. Perfusion and washout of indo-1 decreased LVP by approximately 25% (Fig. 1); this effect is due to the vehicle and to intracellular Ca2+ buffering by indo-1 (6). MnCl2 quenching increased LVP to approximately 85% of preloading values; this is likely due to reversal of indo-1 buffering of Ca2+ by substitution with Mn2+. LVP and m[Ca2+] remained stable during the experiments (Fig. 1).
All analog signals were digitized (PowerLab/16 SP; ADInstruments, Castle Hills, Australia) and recorded at 200 Hz (Chart & Scope Version 3.6.3; ADInstruments) on a Power Macintosh® G4 (Apple Computer, Inc., Cupertino, CA) for later analysis with MATLAB® (The MathWorks, Natick, MA) and Microsoft Excel® (Microsoft Corp., Redmond, WA) software. All variables were averaged over the sampling period of 2.5 s.
After 95 min for equilibration, loading, washout, and MnCl2 perfusion, each protocol lasted 200 min, beginning at 0 min (Fig. 2). Hearts were assigned randomly to one of four different groups: 1) untreated ischemic control hearts (CON, n = 10) were not subjected to preconditioning or given 5-HD; 2) preconditioned hearts (APC, n = 10) were exposed to 1.2 mM (8.8 vol%) sevoflurane for 15 min, followed by a 30-min washout period before ischemia; 3) APC + 5-HD hearts (n = 10) were perfused with 200 μM 5-HD from 5 min before to 15 min after sevoflurane exposure, and this was followed by a 15-min washout of 5-HD before ischemia; and 4) 5-HD hearts (n = 10) received 200 μM 5-HD for 35 min, followed by a 15-min washout before ischemia. There was no difference in sevoflurane concentration between APC and APC + 5-HD hearts. Sevoflurane was almost undetectable (0.05 ± 0.01 mM) in the effluent 30 min after its washout before ischemia. All hearts underwent 30 min of global no-flow ischemia by clamping the aortic inflow; this was followed by 120 min of reperfusion.
All data are mean ± sem. Within groups, data over time for a given variable were compared with a preischemic control period (at 0 min) by Duncan’s comparison of means test whenever univariate analysis of variance F values (P < 0.05) for repeated measures were significant (Super ANOVA 1.11® software for Macintosh; Abacus Concepts, Inc., Berkeley, CA). Among groups, data were compared by analysis of variance at selected time points: before (at 0 and 5 min), during (at 20 min), and after (at 35 and 50 min) sevoflurane exposure; during ischemia (at 80 min); and during reperfusion (at 85 and 200 min). If F values (P < 0.05) were significant, Student-Newman-Keuls post hoc tests were used to compare the four groups. Differences among means were considered significant when P < 0.05. Groups tested were *CON versus APC, †CON versus APC + 5-HD, ‡CON versus 5-HD, §APC versus APC + 5-HD, #APC versus 5-HD, and ¶APC + 5-HD versus 5-HD.
As shown in Figure 3, m[Ca2+] increased continuously during ischemia in each group. This increase was less in APC hearts than in other groups. Perfusion with 5-HD during sevoflurane exposure abrogated this difference, whereas 5-HD alone caused no difference compared with CON.
Figure 4A shows that sevoflurane exposure decreased sysLVP, which was completely reversed with washout. 5-HD given with sevoflurane did not prevent this decrease. APC hearts exhibited better recovery of sysLVP than CON hearts. Developed LVP (data not displayed) was improved throughout reperfusion in APC hearts compared with hearts of other groups (P < 0.05).
Contractility (dLVP/dtmax) and relaxation (dLVP/dtmin) were reversibly decreased during sevoflurane exposure (Fig. 4B); this was not blocked by 5-HD. During reperfusion, contraction and relaxation were better in APC hearts than in other groups.
Sevoflurane reversibly altered heart rate from 245 ± 4 bpm in CON hearts to 181 ± 14*# bpm in APC hearts. This was not blocked by 5-HD (191 ± 14†¶ bpm). Heart rate recovered fully on washout of sevoflurane before ischemia. 5-HD alone caused no difference compared with CON (243 ± 7 bpm). At all other perfusion periods, there was no difference in heart rate among the four groups (data not displayed).
Coronary flow (CF; mL · min−1 · g−1) was not different among groups before ischemia (average, 6.8 ± 0.0). Sevoflurane exposure and 5-HD perfusion caused no significant change in CF. At early reperfusion, each group exhibited a postischemic flow response. CF was significantly higher in the APC group than in the other groups at 5 min of reperfusion: 4.6 ± 0.2* for CON, 6.0 ± 0.3 for APC, 5.0 ± 0.3§ for APC + 5-HD, and 5.2 ± 0.2# for 5-HD. Throughout reperfusion, CF recovered better in the APC group than in the other groups but did not reach preischemic values at 120 min of reperfusion, when CF was 4.7 ± 0.2* for CON, 5.6 ± 0.2 for APC, 4.8 ± 0.3§ for APC + 5-HD, and 4.7 ± 0.2# for 5-HD.
Infarct size (percentage of total ventricular tissue) was decreased significantly only in APC hearts. This decrease was blocked by 5-HD, which, by itself, had no effect on infarct size: 54.9% ± 1.8%* for CON, 31.1% ± 2.0% for APC, 53.0% ± 2.6%§ for APC + 5-HD, and 54.3% ± 2.3%# for 5-HD.
This is the first study to indicate that APC triggered by sevoflurane exposure before global ischemia is associated with attenuated mCa2+ overload during IR and that this effect on mCa2+ overload is reversed by 5-HD, a putative blocker of mKATP channels. As in our previous report on mNADH (7), this study supports the thesis that APC initiates a sequence of events characterized by observable changes in mitochondrial function during ischemia. The mKATP channel opening may be an important factor in this sequence.
Myocardial preconditioning is a phenomenon whereby brief periods of ischemia (IPC) (3) or brief administration of a variety of drugs, (20) including volatile anesthetics (4–7), before prolonged ischemia trigger mechanisms that protect the myocardium against subsequent IR injury. Mitochondria likely play a central role as both triggers and effectors of preconditioning (7,21,22).
Volatile anesthetics are thought to trigger APC, in part, by opening KATP channels (4,5,8). KATP channels are abundant on mitochondrial inner membranes (23,24). The putative mKATP channel opener diazoxide mimics IPC on infarct size (8) and attenuates IR injury (21,25). The opening of mKATP and subsequent K+ influx leads to mitochondrial membrane depolarization, matrix swelling, slowing of ATP synthesis, and accelerated respiration (26). Sufficient mitochondrial membrane depolarization may reduce the driving force for Ca2+ influx through the Ca2+ uniporter (27), thus attenuating mCa2+ overload during ischemia and on reperfusion (14). This is supported by the findings that diazoxide attenuated mCa2+ overload in isolated mitochondria (28), myocytes (29), and isolated rat hearts (13). IPC, as APC, leads to decreased cytosolic Ca2+(6) and mCa2+ overload (13) and to similar changes in NADH (7), which suggests common pathways elicited by APC and IPC.
Our study demonstrates that the APC-induced preservation of function and tissue viability and the attenuated mCa2+ overload are reversed by the concomitant administration of 5-HD. This suggests an important link between mKATP channel function and mCa2+ handling in the cardioprotection by APC and indicates attenuated mCa2+ overload as an observable marker for reduced IR injury. Because IPC and APC both cause a decrease in cytosolic Ca2+ overload during IR (6), lower mCa2+ loading may also result in part from a lower influx of cytosolic Ca2+.
Sevoflurane exposure before ischemia caused a marked but fully reversible reduction in contractility and heart rate; however, this temporary cardiac depression per se cannot be responsible for APC because it was not blocked by 5-HD. It is interesting that despite the temporary cardiac depression, mCa2+ did not change during sevoflurane exposure. Diazoxide perfusion also does not appear to alter mCa2+ under normoxic conditions (13). Therefore, sevoflurane exposure without ischemia may not open mKATP channels directly but rather may “prime”(30) them directly or indirectly to open faster and/or more during subsequent ischemia. Volatile anesthetics, as well as diazoxide, cause small but significant changes in mitochondrial energetics (7,31). Furthermore, we have demonstrated a role for APC in attenuating mNADH accumulation during ischemia (7). Thus, the attenuation of mCa2+ by APC may be the result of a partial preservation of mitochondrial bioenergetics rather than their cause.
The higher mCa2+ signals on reperfusion for each group persisted until approximately 30 minutes of reperfusion and then reached preischemic values during the last 90 minutes of reperfusion. Thus, we did not observe a continuously increased mCa2+ signal or an attenuated mCa2+ signal in preconditioned hearts, as was reported with shorter reperfusion protocols in intact rat hearts (13) or isolated rabbit ventricular myocytes (29). Infarction in our model was concentric and subepicardial; thus, infarcted cells, as well as viable cells, underlie the fiberoptic probe. The overall F measured in our study is therefore likely a product of average mCa2+ signals per cell and the number of viable cells contributing to the observed F. We suggest that lower mCa2+ in a larger number of viable cells in preconditioned hearts may balance out higher mCa2+ in a smaller number of viable cells in nonpreconditioned hearts so that the overall F signal during reperfusion is similar among groups. If this is so, the mCa2+ measurements obtained on reperfusion must be interpreted in context with the fraction of viable cells on reperfusion, i.e., the percentage of noninfarcted tissue.
Possible limitations of our study are that the use of a high-Po2, blood-free perfusate causes a greater use of CF reserve and that volatile anesthetics may modify neutrophil function related to IR injury in in vivo hearts. Moreover, the isolated heart is devoid of autonomic and hormonal input, which could influence IR injury. Although many investigators use the rat heart as a model for APC, the guinea pig is closer in design to the human heart, as we (2,6) and others (32–34) have observed. We used only one pulse of a large concentration of sevoflurane to induce APC. However, from our studies in the same preparation, we know that APC can also be induced with two pulses of smaller concentrations (5,6) and that there is a concentration-dependent effect of APC induced by sevoflurane (7). It is important to note that conclusions on the role of mKATP channel opening in APC depend on the selectivity of 5-HD as a blocker of mKATP channels.
In summary, we have shown that APC not only afforded cardioprotection against mechanical dysfunction and tissue necrosis on reperfusion after myocardial ischemia, but also attenuated mCa2+ overload during ischemia and on reperfusion. Reversal of these effects by the putative mKATP channel blocker 5-HD suggests that direct or indirect mKATP channel opening is part of the triggering mechanism of APC. This study extends our knowledge of the triggering mechanism of volatile anesthetics to protect against myocardial IR injury. APC could prove advantageous in patients undergoing surgery who are at risk for cardiac ischemia.
The authors thank James Heisner for technical assistance and Dr. Srinivasan Varadarajan, Dr. Ming Tao Jiang, Dr. Jianzong An, Anita Tredeau, and Mary Ziebell for their valuable contributions to this study.
1. Warltier DC, al-Wathiqui MH, Kampine JP, Schmeling WT. Recovery of contractile function of stunned myocardium in chronically instrumented dogs is enhanced by halothane or isoflurane. Anesthesiology 1988; 69: 552–65.
2. Varadarajan SG, An JZ, Novalija E, Stowe DF. Sevoflurane before or after ischemia improves contractile and metabolic function while reducing myoplasmic Ca2+
loading in intact hearts. Anesthesiology 2002; 96: 125–33.
3. Murry CE, Jennings RB, Reimer KA. Preconditioning with ischemia: a delay of lethal cell injury in ischemic myocardium. Circulation 1986; 74: 1124–36.
4. Kersten JR, Schmeling TJ, Pagel PS, et al. Isoflurane mimics ischemic preconditioning via activation of KATP
channels: reduction of myocardial infarct size with an acute memory phase. Anesthesiology 1997; 87: 361–70.
5. Novalija E, Fujita S, Kampine JP, Stowe DF. Sevoflurane mimics ischemic preconditioning effects on coronary flow and nitric oxide release in isolated hearts. Anesthesiology 1999; 91: 701–12.
6. An JZ, Varadarajan SG, Novalija E, Stowe DF. Ischemic and anesthetic preconditioning reduces cytosolic [Ca2+
] and improves Ca2+
responses in intact hearts. Am J Physiol Heart Circ Physiol 2001; 281: H1508–23.
7. Riess ML, Camara AK, Chen Q, et al. Altered NADH and improved function by anesthetic and ischemic preconditioning in guinea pig intact hearts. Am J Physiol Heart Circ Physiol 2002; 283: H53–60.
8. Garlid KD, Paucek P, Yarov-Yarovoy V, et al. Cardioprotective effect of diazoxide and its interaction with mitochondrial ATP-sensitive K+
channels: possible mechanism of cardioprotection. Circ Res 1997; 81: 1072–82.
9. Hansford RG. Relation between mitochondrial calcium transport and control of energy metabolism. Rev Physiol Biochem Pharmacol 1985; 102: 1–72.
10. Miyamae M, Camacho SA, Weiner MW, Figueredo VM. Attenuation of postischemic reperfusion injury is related to prevention of [Ca2+
]m overload in rat hearts. Am J Physiol 1996; 271: H2145–53.
11. Di Lisa F, Bernardi P. Mitochondrial function as a determinant of recovery or death in cell response to injury. Mol Cell Biochem 1998; 184: 379–91.
12. Varadarajan SG, An JZ, Novalija E, et al. Changes in [Na+
]i, compartmental [Ca2+
], and NADH with dysfunction after global ischemia in intact hearts. Am J Physiol Heart Circ Physiol 2001; 280: H280–93.
13. Wang L, Cherednichenko G, Hernandex L, et al. Preconditioning limits mitochondrial Ca2+
during ischemia in rat hearts: role of KATP
channels. Am J Physiol Heart Circ Physiol 2001; 280: H2320–8.
14. Ferrari R. The role of mitochondria in ischemic heart disease. J Cardiovasc Pharmacol 1996; 28 (Suppl 1): S1–10.
15. Epstein RH, Stein AL, Marr AT, Lessin JB. High concentration versus incremental induction of anesthesia with sevoflurane in children: a comparison of induction times, vital signs, and complications. J Clin Anesth 1998; 10: 41–5.
16. Grynkiewicz G, Poenie M, Tsien RY. A new generation of Ca2+
indicators with greatly improved fluorescence properties. J Biol Chem 1985; 260: 3440–50.
17. Brandes R, Figueredo VM, Camacho SA, et al. Quantitation of cytosolic [Ca2+
] in whole perfused rat hearts using indo-1 fluorometry. Biophys J 1993; 65: 1973–82.
18. Brandes R, Figueredo VM, Camacho SA, et al. Investigation of factors affecting fluorometric quantitation of cytosolic [Ca2+
] in perfused hearts. Biophys J 1993; 65: 1983–93.
19. Chance B, Williamson JR, Jamieson D, Schoenner B. Properties and kinetics of reduced pyridine nucleotide fluorescence of the isolated and in vivo rat heart. Biochem Z 1965; 341: 357–77.
20. Okubo S, Xi L, Bernardo NL, et al. Myocardial preconditioning: basic concepts and potential mechanisms. Mol Cell Biochem 1999; 196: 3–12.
21. Gross GJ, Fryer RM. Sarcolemmal versus mitochondrial ATP-sensitive K+
channels and myocardial preconditioning. Circ Res 1999; 84: 973–9.
22. Novalija E, Varadarajan SG, Camara AKS, et al. Anesthetic preconditioning: triggering role of reactive oxygen and nitrogen species in isolated hearts. Am J Physiol Heart Circ Physiol 2002; 283: H44–52.
23. Inoue I, Nagase H, Kishi K, Higuti T. ATP-sensitive K+
channel in the mitochondrial inner membrane. Nature 1991; 352: 244–7.
24. Paucek P, Mironova G, Mahdi F, et al. Reconstitution and partial purification of the glibenclamide-sensitive, ATP-dependent K+
channel from rat liver and beef heart mitochondria. J Biol Chem 1992; 267: 26062–9.
25. Sato T, Marban E. The role of mitochondrial KATP
channels in cardioprotection. Basic Res Cardiol 2000; 95: 285–9.
26. Holmuhamedov EL, Jovanovic S, Dzeja PP, et al. Mitochondrial ATP-sensitive K+
channels modulate cardiac mitochondrial function. Am J Physiol 1998; 275: H1567–76.
27. Gunter TE, Gunter KK, Sheu SS, Gavin CE. Mitochondrial calcium transport: physiological and pathological relevance. Am J Physiol 1994; 267: C313–39.
28. Holmuhamedov EL, Wang L, Terzic A. ATP-sensitive K+
channel openers prevent Ca2+
overload in rat cardiac mitochondria. J Physiol 1999; 519 (Pt 2):347–60.
29. Murata M, Akao M, O’Rourke B, Marban E. Mitochondrial ATP-sensitive potassium channels attenuate matrix Ca2+
overload during simulated ischemia and reperfusion: possible mechanism of cardioprotection. Circ Res 2001; 89: 891–8.
30. Zaugg M, Lucchinetti E, Spahn DR, et al. Volatile anesthetics mimic cardiac preconditioning by priming the activation of mitochondrial KATP
channels via multiple signaling pathways. Anesthesiology 2002; 97: 4–14.
31. Liu Y, Sato T, O’Rourke B, Marban E. Mitochondrial ATP-dependent potassium channels: novel effectors of cardioprotection? Circulation 1998; 97: 2463–9.
32. Steinfath M, Chen YY, Lavicky J, et al. Cardiac alpha 1-adrenoceptor densities in different mammalian species. Br J Pharmacol 1992; 107: 185–8.
33. Kanai A, Salama G. Optical mapping reveals that repolarization spreads anisotropically and is guided by fiber orientation in guinea pig hearts. Circ Res 1995; 77: 784–802.
34. Takeishi Y, Huang Q, Wang T, et al. Src family kinase and adenosine differentially regulate multiple MAP kinases in ischemic myocardium: modulation of MAP kinases activation by ischemic preconditioning. J Mol Cell Cardiol 2001; 33: 1989–2005.