INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS AND MATERIALS RESULTS DISCUSSION REFERENCES

ISCHEMIC PRECONDITIONING (IPC) is a well-known phenomenon in which brief episodes of ischemia and reperfusion paradoxically protect the heart against subsequent lethal ischemia (22). ATP-dependent potassium (K_ATP) channels have been proposed to be the end-effector of this protection (13). The protection by K_ATP channels was initially thought to be via the surface membrane channel (sK_ATP) of myocytes. However, recent studies have shown that K_ATP channels in the mitochondrial inner membrane (mK_ATP) are responsible for the protection (12, 13, 16, 17). In general, the following signal transduction pathways of IPC have been supported by a majority of studies: during brief ischemia, a precondition state is triggered by various mechanisms (trigger; such as activation of adenosine and ₁-adrenergic receptors and elevation of intracellular Ca²⁺), which then cause translocation/activation of protein kinase C (PKC) and probably other downstream kinases. During the prolonged ischemia, the mK_ATP channel (mediator) is phosphorylated and thus opens earlier and/or more to mediate the protection by unknown mechanisms.

Several recent studies have suggested that mK_ATP channels can trigger cardioprotection as well. Using Ca²⁺ paradox (Ca²⁺ depletion followed by repletion) to injure rat myocardium, Wang et al. (31) have shown that diazoxide, a selective mK_ATP channel opener in myocardium, administrated during Ca²⁺ paradox improved functional recovery. Interestingly, hearts treated with diazoxide but followed by a period of washout before the Ca²⁺ paradox were also protected. These results indicate that mK_ATP opening can mediate and trigger cardioprotection. Coadministration of 5-hydroxydecanoic acid (5-HD), a selective mK_ATP channel blocker, or PKC inhibitors blocked the protection from diazoxide. The role of mK_ATP channels as a trigger was supported by a study from Pain et al. (25) in which 5 min of diazoxide treatment reduced myocardial infarction even when diazoxide was washed out for up to 30 min before the 30-min lethal ischemia. 5-HD blocked the protection from diazoxide when they were coadministered. However, 5-HD failed to block the protection when it was given only during the lethal ischemia. The study from Pain et al. (25) suggests that mK_ATP channels may be only a trigger, but not a mediator, of protection.

It is very important to clarify the roles of mK_ATP channels in cardioprotection. Although 5-HD is now widely used as a selective mK_ATP channel blocker in cardiac myocytes, it has a low potency. Previous studies have indicated that some ischemia-related changes may decrease the potency of K_ATP channel inhibitors. It has been shown that a potent K_ATP channel blocker, glybenclamide, becomes less effective during metabolic inhibition (10). Sato et al. (28) have shown that up to 2 mM of 5-HD (compared with 200 µM in untreated cells; see Ref. 16) is needed to inhibit diazoxide-induced mK_ATP channel opening in phorbol 12-myristate 13-acetate (PMA)-treated myocytes (28), suggesting that PKC phosphorylation of the channel may make the channel more resistant to 5-HD. In this study, we further investigate the roles of the mK_ATP channel as a trigger and a mediator. We used a model of myocardial infarction produced by 30 min of regional ischemia and 2 h of reperfusion in isolated rabbit hearts. We used various concentrations of 5-HD and glybenclamide to block K_ATP channels. We confirmed that mK_ATP can be a trigger for protection. Our results suggest that protection may be induced by an increase of intracellular Ca²⁺ and subsequent activation of PKC. We also showed that inhibition of mK_ATP channels during lethal ischemia also blocked the protection of diazoxide, indicating a role as a mediator as well.

	METHODS AND MATERIALS

TOP ABSTRACT INTRODUCTION METHODS AND MATERIALS RESULTS DISCUSSION REFERENCES

The present study was conducted in accordance with the Guide for the Care And Use of Laboratory Animals published by the National Research Council (1996, Washington, DC) and was approved by the Institutional Animal Care and Use Committee of Otsuka America Pharmaceutical.

Chemicals

Myocardial Infarction Studies

Myocardial infarction. Male New Zealand White rabbits, weighing 1.5-2.5 kg, were used in this study. Rabbits were anesthetized with pentobarbital sodium (30 mg/kg) through a marginal ear vein, and the animals were then mechanically ventilated by a tracheal cannula with room air supplemented with oxygen. The heart was exposed through a left thoracotomy in the fourth intercostal space, and the pericardium was opened. A snare was placed around a major branch of the left coronary artery of the animals using a 4-0 suture. The suture ends were passed through a small segment of pliable polyethylene tubing to form a snare. The heart was quickly excised by an incision at the base of the heart and was put in ice-cold Krebs-Henseleit bicarbonate buffer. The heart was then attached to a Langendorff apparatus by the aortic root and was perfused with nonrecirculating modified Krebs-Henseleit buffer [composed of (in mM) 118 NaCl, 25 NaHCO₃, 1.2 KH₂PO₄, 4.75 KCl, 1.2 MgSO₄, 2 CaCl₂, and 10 dextrose] at a constant pressure of 75 mmHg. The perfusate was bubbled with a 95% O₂-5% CO₂ gas mixture, and the bubbling rate was adjusted to maintain physiological pH (7.35-7.45). Perfusate temperature was maintained at 38°C by a circulating water jacket surrounding the buffer reservoirs. The heart was also maintained at 38°C via a water-jacketed housing in which it was suspended. The open top of the jacket was covered with a piece of Parafilm to maintain the humidity and temperature. The pulmonary artery was cannulated for coronary flow (CF) rate measurement. A saline-filled latex balloon, connected via a catheter to a pressure transducer (Ohmeda), was inserted in the left ventricle and inflated to yield an end-diastolic pressure of 0-5 mmHg. The pressure transducer was connected to a Grass Chart Recorder (model 7) to record left ventricular pressure and its first derivative (dP/dt) and heart rate. CF was measured by a timed collection of the effluent in a graduated cylinder. Hearts with left ventricular developing pressure <85 mmHg at the end of the 20-min equilibrium period were not included in the study. Regional ischemia was performed by pulling the silk tightly through the tubing and clamping the tube with a hemostat. All of the hearts had a 30-min ischemia and 2-h reperfusion. Reperfusion was realized by releasing the snare. Myocardial ischemia and reperfusion were confirmed by a decrease of left ventricular developed pressure (LVDP) and CF and partial recovery of these two parameters, respectively.

Protocols. The protocols are summarized in Fig. 1. To study the role of mK_ATP channels as a trigger, we performed experiments on eight groups of hearts. The control group had a 30-min ischemia and 2-h reperfusion only (without any drug treatment). The Diaz(E) group had 5 min of 100 µM diazoxide perfusion followed by a 10-min drug-free period before the 30 min of ischemia. The Nif(E), Che(E), and 5-HD(E) groups had 10 min of 100 nM nifedipine, 5 µM chelerythrine, or 50 µM 5-HD, respectively, starting 20 min before the 30 min of ischemia. Hearts in the Nif(E) + Diaz(E), Che(E) + Diaz(E), and 5-HD(E) + Diaz(E) groups were perfused with the same treatment as in the Nif(E), Che(E), and 5-HD(E) groups, respectively, in addition to the same treatment in the Diaz(E) group.

View larger version (14K): [in this window] [in a new window]   Fig. 1.   Experimental protocol. The study was separated into two parts: one part to study the trigger phase and the other part to study the mediator phase. The x-axis shows the start of the 30-min ischemia, which was designated as time 0. Shaded bars, period of drug (labeled inside the bar) perfusion; E, drugs administrated only before the 30 min ischemia; L, drugs given during ischemia; Diaz, diazoxide; Nif, nifedipine; Che, chelerythrine; 5-HD, 5-hydroxydecanoic acid sodium; 5-HD(hL), high concentration (200 µM) of 5-HD; Gly, glybenclamide.

Cellular Studies

L-type Ca²⁺ current and flavoprotein fluorescence measurement. The L-type Ca²⁺ currents and flavoprotein fluorescence were measured similarly as reported by Liu et al. (16). Flavoprotein fluorescence, reflecting the redox state of the mitochondria, was used as an index of the mK_ATP channel opening (16). Ventricular myocytes were isolated from adult rabbit hearts by conventional enzymatic dissociation (15). Cells were cultured on laminin-coated coverslips in M199 culture medium with 5% FBS at 37°C. Experiments were performed on the next day. The cells were perfused with a modified Tyrode solution consisting of (in mM) 140 NaCl, 5 KCl, 2 CaCl₂, 1 MgCl₂, and 10 HEPES (pH 7.4 with NaOH). For whole cell patch recordings, the internal pipette solution contained (in mM) 120 potassium glutamate, 25 KCl, 0.5 MgCl₂, 10 potassium EGTA, 10 HEPES, and 1 MgATP (pH 7.2 with KOH). Whole cell currents were elicited every 6 s from a holding potential of 80 mV by two consecutive steps to 40 mV (for 100 ms) to inactivate Na⁺ channels and 0 mV (for 380 ms) for L-type Ca²⁺ and K_ATP current measurements. The peak inward currents at 0 mV were taken as L-type Ca²⁺ currents, and K_ATP currents were measured 200 ms into the pulse. Endogenous flavoprotein fluorescence was excited using a xenon arc lamp with a bandpass filter centered at 480 nm, but only during the 100-ms step to 40 mV to minimize photobleaching. Emitted fluorescence was recorded at 530 nm by a photomultiplier tube and was digitized (Digidata 1200; Axon Instruments). Relative fluorescence was averaged during the excitation window and was calibrated using the values after dinitrophenol and sodium cyanide exposure. Experiments were performed at room temperature.

Intracellular Ca²⁺ measurement. We used indo 1-AM to measure intracellular Ca²⁺ as reported by Bassani et al. (3). Freshly isolated rabbit ventricular myocytes were loaded with indo 1 by incubation with 5 µM indo 1-AM for 15 min at room temperature. After being loaded, cells were washed with Tyrode solution extensively and were allowed to set for 1 h for intracellular indo 1 deesterification. The cells were then superfused with Tyrode solution and field stimulated (square wave, 0.5 Hz) through a pair of platinum electrodes. Excitation wavelength was 365 nm, and fluorescence emitted by the cell at 405 and 485 nm was recorded and digitized at 2 kHz. The fluorescence ratio of 405 to 485 nm was used as a measure of intracellular Ca²⁺ levels. To obtain the peak ratios, we averaged 10 sequential recordings and then smoothed further by averaging the adjacent 10 points using Microcal Origin software (Microcal Software, Northampton, MA). Peak values were derived from the averaged, smoothed traces. Experiments were performed at room temperature.

Statistical analysis. Data are presented as means ± SE, and statistics were performed using SigmaStat (Jandel, San Rafael, CA). ANOVA combined with Tukey's post hoc test were used to test for differences among groups for infarct size. Hemodynamic data were analyzed by one-way ANOVA with repeated measurements combined with Tukey's post hoc test within the group. Fluorescence data were analyzed by a one-way ANOVA combined with Tukey's post hoc test. P < 0.05 was considered significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS AND MATERIALS RESULTS DISCUSSION REFERENCES

The cardiac functional changes are summarized in Tables 1 and 2. Diazoxide (100 µM) did not significantly affect LVDP, maximal dP/dt (dP/dt_max), or heart rate. However, it did significantly increase CF. Nifedipine (100 nM) reduced LVDP and dP/dt_max and increased CF. 5-HD did not affect any of the cardiac function indexes and had no effect on the diazoxide-induced CF increases. The only significant effect from chelerythrine was to increase CF. Regional ischemia significantly reduced LVDP, dP/dt_max, and CF in all of the groups, and reperfusion caused a partial cardiac function recovery.

The average sizes of risk area and infarct area are summarized in Table 3. There is no significant difference in risk size among the groups. The myocardial infarct size data for the trigger phase are shown in Fig. 2. Pretreatment with diazoxide before the 30 min of ischemia [Diaz(E) group] caused a 75% reduction of infarct size (to 9 ± 3 from 35 ± 3% in the controls). This protection was blocked by the coadministration of 100 nM nifedipine, 5 µM chelerythrine, or 50 µM 5-HD. Nifedipine, chelerythrine, or 5-HD alone did not significantly affect infarct size. The infarct sizes for the mediator phase are summarized in Fig. 3. Administration of 50 µM of 5-HD during the 30 min of ischemia only did not block the early treatment protection of diazoxide. However, increasing the concentration of 5-HD to 200 µM completely eliminated the protection of diazoxide [29 ± 3% of infarction in the Diaz(E) + 5-HD(hL) group]. Glybenclamide (10 µM) given during the 30 min of ischemia also blocked the protection (34 ± 3% infarction). Diazoxide given throughout the 30 min of ischemia [Diaz(L) group] reduced infarct size (6 ± 2% infarction), and this protection was not blocked by coadministration with chelerythrine (7 ± 1% infarction).

View larger version (16K): [in this window] [in a new window]   Fig. 2.   Infarct size data of the groups in the trigger phase. , Infarct size of individual hearts; , averages. Note that a brief treatment with diazoxide before the 30 min of ischemia significantly reduced infarction. The protection was blocked by Nif, Che, and 5-HD. See text for group explanations.

View larger version (18K): [in this window] [in a new window]   Fig. 3.   Infarct size data of the groups in the mediator phase. , Infarct size of individual hearts; , averages. Note that 5-HD at 50 µM did not block the protection, but 5-HD at 200 µM did. Che did not block this protection but Gly did. See Fig. 1 group explanations.

We investigated whether diazoxide affects L-type Ca²⁺ currents and whether nifedipine directly affects mK_ATP channel opening by diazoxide by simultaneously monitoring L-type Ca²⁺ channel currents and mitochondrial oxidation as an index of mK_ATP channel opening (16). Figure 4A shows representative time courses of mK_ATP opening, L-type Ca²⁺ current, and sK_ATP currents from one cell. In this cell, diazoxide (100 µM) induced 55% mitochondrial oxidation, an indication of mK_ATP channel opening. Diazoxide had no effect on L-type Ca²⁺ or sK_ATP currents. Nifedipine (100 nM) did not alter the effect of diazoxide on mK_ATP channel opening, but it inhibited >80% of the L-type Ca²⁺ currents. Nifedipine did not have an effect on sK_ATP channels. Figure 4B summarizes the changes of mK_ATP channels and L-type Ca²⁺ currents from six cells. We also studied whether blocking PKC would affect the ability of diazoxide to open mK_ATP channels. We could not use chelerythrine as a blocker in this part of the study because it is also green fluorescent, but neither staurosporine, a nonselective protein kinase inhibitor, nor polymyxin B, a selective PKC inhibitor, affected the mitochondrial oxidation induced by diazoxide (data not shown), suggesting that PKC activity is not required for diazoxide to open mK_ATP channels.

View larger version (25K): [in this window] [in a new window]   Fig. 4.   Effects of diazoxide and nifedipine on mitochondrial oxidation, sarcolemmal L-type Ca2+ current, and KATP currents (IK,ATP). A: representative recordings from one cell. Diazoxide increased mitochondrial oxidation, and this effect was not affected by nifedipine. Although diazoxide did not affect L-type Ca2+ currents, they were blocked largely by nifedipine. Neither diazoxide nor nifedipine affected surface KATP currents. DNP, dinitrophenol. B: data on mitochondrial oxidation and L-type Ca2+ currents.

We also studied the effect of diazoxide on intracellular Ca²⁺ in electric field-stimulated ventricular myocytes (n = 6). As shown in Fig. 5, diazoxide (100 µM) significantly increased the peak values of the indo 1 ratio (405/485 nm), indicating that it elevates peak intracellular Ca²⁺ levels during a cell contraction, although the increase is relatively small (7%). 5-HD (100 µM) blocked the effect of diazoxide.

View larger version (18K): [in this window] [in a new window]   Fig. 5.   Effect of diazoxide on intracellular Ca2+ during a cell contraction. Intracellular Ca2+ was measured by the ratio (405/485 nm) of indo 1 fluorescence. A: representative recordings from one cell. B: summarized data.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS AND MATERIALS RESULTS DISCUSSION REFERENCES

Strong evidence has supported the role of K_ATP channels as mediators in IPC and cardioprotection against ischemia and reperfusion injury (13). IPC is triggered by various receptor activation and subsequent PKC and downstream kinase stimulations. Phosphorylation of K_ATP channels primes the channel to open earlier and/or more during the lethal ischemia to confer protection (13). Although the sK_ATP channel was the candidate earlier on, later pharmacological studies have shown that the mK_ATP channel is responsible for protecting the heart (12, 16); diazoxide, a selective mK_ATP channel opener in cardiac myocytes, reduces myocardial injury, and 5-HD, a selective mK_ATP channel inhibitor, eliminates protection from diazoxide and IPC (16). Interestingly, recent studies have suggested that mK_ATP channel opening can trigger cardioprotection. Wang et al. (31) showed that rat hearts treated with diazoxide but followed by a period of washout improved functional recovery after ischemia and reperfusion in rat hearts. The protection can be blocked by 5-HD, chelerythrine, and nifedipine, coadministered with diazoxide. Consistent with the role of mK_ATP channel opening as a trigger, Pain et al. (25) showed that 5 min of diazoxide pretreatment reduced myocardial infarction even when diazoxide was washed out for up to 30 min before the lethal ischemia in isolated rabbit hearts. 5-HD abolished the protection from diazoxide when they were coadministrated. Consistent with these findings, we showed in this study that diazoxide can trigger the heart into a preconditioned state even after diazoxide has been washed out. We also showed that this protection can be blocked by coadministration of 5-HD, chelerythrine, and nifedipine. Although 5-HD blocked the opening of mK_ATP channels by diazoxide, nifedipine had no direct effect.

The blockade by 5-HD suggests that the trigger and memory phase of protection is mediated by opening of mK_ATP channels. How opening of mK_ATP channels can put the heart into a preconditioned state is not clear. The blocking by nifedipine, as shown by Wang et al. (30, 31) and by this study, indicates a role of L-type Ca²⁺ channels in the trigger phase of protection by mK_ATP channels. A transient increase of intracellular Ca²⁺ has been shown to precondition rat and canine hearts (20, 21, 26), possibly by a PKC-mediated mechanism (21). Recently, Cain et al. (6) demonstrated that clinical L-type Ca²⁺ blockers block IPC in human atrial trabeculae. This evidence has implicated Ca²⁺ entry through the L-type Ca²⁺ channel as an effective trigger of preconditioning. Our results show that diazoxide does not affect L-type Ca²⁺ currents in rabbit ventricular myocytes. When intracellular Ca²⁺ levels were measured using indo 1 Ca²⁺ dye, however, we did demonstrate a significant elevation of peak intracellular Ca²⁺ by diazoxide during cell contraction. The resting level of Ca²⁺ was not changed significantly by diazoxide. We followed the method of indo 1 loading described by Bassani et al. (3) in which they confirmed that this loading protocol did not cause significant compartmentalization. Thus the change we saw should reflect cytosolic Ca²⁺.

Mitochondria have a high capacity for Ca²⁺ buffering and have been shown to play a role in regulating cytosolic Ca²⁺ in neurons and cardiac myocytes (1, 9). Although it is still controversial (34), rapid mitochondrial Ca²⁺ transients have been demonstrated during excitation-contraction in adult rabbit cardiac myocytes (24); mitochondria take up Ca²⁺ during systole and release Ca²⁺ during diastole. Mitochondrial Ca²⁺ influx is driven primarily by the mitochondrial inner membrane potential. Thus the driving force for Ca²⁺ movement will favor uptake into the matrix through a Ca²⁺ uniporter once the cytosolic concentration of Ca²⁺ close to the mitochondria rises such as during a systole. During diastole, mitochondria release Ca²⁺, presumably through an Na⁺/Ca²⁺ exchanger (9). Opening of mK_ATP channels leads to potassium influx in mitochondria and would tend to dissipate the membrane potential established by the proton pump (14, 16). A mitochondrial membrane depolarization would then cause Ca²⁺ release from mitochondria, as shown in neuronal cells (1, 9). In isolated cardiac mitochondria, Holmuhamedov et al. (14) have shown that K_ATP channel openers depolarized the mitochondrial membrane and released accumulated Ca²⁺ in mitochondria. We postulate that the elevated peak cytosolic Ca²⁺ concentration during a contraction in diazoxide-treated cells is caused by the activation of mK_ATP channels and subsequent mitochondrial membrane depolarization. During systole when cytosolic Ca²⁺ is elevated, the depolarized mitochondria would uptake less Ca²⁺, thus resulting in higher systolic Ca²⁺ levels. A brief elevation of cytosolic Ca²⁺ then induces protection. This is also consistent with the observations that a brief period of elevation of intracellular Ca²⁺ [by Ca²⁺ depletion and repletion (19), by increase of extracellular Ca²⁺ (26), or by -adrenergic receptor activation (20)] can precondition the heart. Transient elevation of intracellular Ca²⁺ is a strong activator of PKC (19, 20), and thus it is not surprising to see a blockade of protection with chelerythrine in this study. It has been shown that PKC isoforms can be translocated to the mitochondria by diazoxide (31). A localized increase of intracellular Ca²⁺ near the mitochondria may be sufficient for PKC activation. This might explain why a small elevation of peak cytosolic Ca²⁺ (~7% increase of baseline value) by diazoxide can provide a significant cardioprotection. A confocal microscopic study will be required to resolve spatiotemporal Ca²⁺ changes by diazoxide.

There is evidence suggesting that PKC activation leads to the activation of p38 mitogen-activated protein kinase (8). Interestingly, Baines et al. (2) have shown that the p38 activator anisomycin reduces infarction, and the protection is blocked by 5-HD. It is not known what the roles of p38 and other kinases are in the trigger and mediator phases of diazoxide, and future studies are needed to address this issue.

That 5-HD and glybenclamide administrated during the lethal ischemia blocked the protection indicates that the mK_ATP channel is a mediator of protection. Our results are consistent with in vivo dog and rat studies from the laboratory of Gross et al. (11, 32). Interestingly, several recent studies emerged to show that blocking mK_ATP channels by 5-HD during the lethal ischemia only (after the preconditioning stimuli) abolished the delayed protection (second window of protection) in in vivo rabbit and rat hearts (4, 5, 23, 29), suggesting that mK_ATP channels are also mediators of the delayed protection. The role as a mediator by the mK_ATP channel in the delayed protection is also demonstrated by Carroll and Yellon (7) in a human cardiomyocyte-derived cell line. However, the results from the above-mentioned studies and ours are in contrast to a study by Pain et al. (25), who showed that 200 µM 5-HD given only during the lethal ischemia did not block protection. The study by Pain et al. (25) and our study employed the same model of isolated rabbit hearts. There are no readily available explanations for such opposing outcomes. Our 5-HD results were confirmed with the nonselective but potent K_ATP channel blocker glybenclamide, whereas Pain et al. (25) used 5-HD alone. Although 5-HD is now widely used as a selective mK_ATP channel blocker in cardiac myocytes, its potency is relatively weak. There is some evidence to suggest that PKC (a major mediator of IPC)-phosphorylated mK_ATP channels become more resistant to 5-HD. Sato et al. (28) have shown that up to 2 mM of 5-HD (compared with 200 µM in untreated cells; see Ref. 16) is needed to inhibit diazoxide-induced mK_ATP channel opening in PMA-treated myocytes. Our results also support that notion: 5-HD at 50 µM blocked the trigger effect of diazoxide, but a much higher concentration of 5-HD (200 µM) is required to block the protection during the lethal ischemia, when the channel probably has been phosphorylated by PKC or other downstream kinases. It has also been shown that glybenclamide becomes less potent during a severe metabolic inhibition (10). Thus it is possible that a higher concentration of 5-HD would be required to block protection in the study by Pain et al. (25).

It has been suggested that the preconditioning renders the mK_ATP channel to open more and/or earlier during a protract ischemia (28). Preconditioning can be accomplished via various PKC-coupled stimulations (8). Our data suggest that the trigger effect of diazoxide is to elevate intracellular Ca²⁺ and subsequent activation of PKC, which then primes the channel. The primed channel then opens more and/or earlier to mediate the protection. If PKC is inhibited during the trigger phase, then the channel will not be primed and protection is lost. However, when mK_ATP channels are directly opened by diazoxide during the lethal ischemia, the priming effect by PKC is obviously not required for the protection. In this situation, a PKC inhibitor will not block the protection of diazoxide, as supported by this study. Our results are also entirely in agreement with a study by Miura et al. (18) and suggest that PKC is upstream of mK_ATP channels and PKC phosphorylation is not required for the protection when the mK_ATP channel is directly opened by a channel opener. Otherwise, PKC activation is required to modulate the channel to open earlier and/or more during lethal ischemia.

Ockaili et al. (23) have recently shown that diazoxide-induced protection is also blocked by a nitric oxide (NO) synthase inhibitor. Although it is currently unknown how the NO pathway fits in the signal transduction scheme of diazoxide, a study by Sasaki et al. (27) may have provided some link. Sasaki et al. showed that, although NO itself activates mK_ATP channels slightly, it greatly potentiates the ability of diazoxide to open these channels. Thus endogenous NO may be an important factor for the mK_ATP channel opening, especially when a subthreshold amount of diazoxide is given.

It is not known how opening of mK_ATP channels can protect myocytes during a lethal ischemia. It has been suggested that mK_ATP channel opening depolarizes the mitochondrial membrane and decreases the mitochondrial Ca²⁺ influx driving force and thus ameliorates mitochondrial Ca²⁺ overload during ischemia (16). Interestingly, a recent study by Ylitalo et al. (33) showed that in isolated rat hearts, IPC accelerated the ischemic mitochondrial membrane potential decrease, consistent with this theory. The overall intracellular Ca²⁺ levels during ischemia and reperfusion were lower in IPC hearts, although Ca²⁺ levels in mitochondria and cytosol were not separated in that study.

In summary, we have shown that mK_ATP channel opening can trigger and mediate cardioprotection in isolated rabbit hearts. The trigger phase may be caused by an increase of intracellular Ca²⁺ and subsequent PKC activation, but the mediator phase is induced by a yet unknown mechanism.