INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

ISCHEMIC PRECONDITIONING (PC) is a phenomenon in which a single period or multiple brief periods of ischemia have been shown to protect the heart against a more prolonged ischemic insult (7, 30, 37). Although it has been proposed that a number of substances and signaling pathways may be involved in mediating the cardioprotective effect of PC (7, 37), the precise mechanism by which PC exerts its protective effects is not known. Recent studies have focused on the potential role of mitochondrial ATP-sensitive potassium (K_ATP) channels as the effectors of protection in PC (13).

Opening of mitochondrial K_ATP channels has been shown to protect the heart during ischemia-reperfusion (11, 14), although the mechanism of this protection is largely unknown. One hypothesis proposed is that opening the mitochondrial K_ATP channel dissipates the inner mitochondrial membrane potential established by the proton pump (26). This dissipation is expected to reduce the driving force for Ca²⁺ influx through the Ca²⁺ uniporter, thus attenuating mitochondrial calcium concentration ([Ca²⁺]_m) overload (8-10). This postulate has been supported in a recent study using isolated rat mitochondria and intact rat cardiomyocytes in which K_ATP channel openers prevented [Ca²⁺]_m overload by both reducing the driving force for Ca²⁺ uptake via the uniporter and by activating cyclosporin-sensitive Ca²⁺ efflux (20).

[Ca²⁺]_m overload has been closely correlated with mitochondrial and myocardial damage and cell death (5, 8, 18, 27, 28, 32). Although limiting [Ca²⁺]_m overload is expected to contribute to cardioprotection, direct evidence that opening mitochondrial K_ATP channel limits ischemia-induced [Ca²⁺]_m overload is lacking.

Recent fluorescent techniques allow the measurement of total intracellular calcium concentration ([Ca²⁺]_i) and [Ca²⁺]_m in beating, perfused hearts by using the calcium indicator indo 1 and cytosolic calcium quenching using manganese (3, 4, 34). With the use of this methodology, Miyamae et al. (27) have shown that pharmacological limitation of [Ca²⁺]_m overload resulted in greater functional recovery after ischemia-reperfusion. These findings, in the face of no change in cytosolic calcium concentration ([Ca²⁺]_c), suggest that limiting [Ca²⁺]_m overload is protective independent of [Ca²⁺]_c.

Thus the purpose of this study was to test the hypotheses that 1) PC limits the accumulation of [Ca²⁺]_m during ischemia and reperfusion, and 2) this effect is mediated by opening mitochondrial K_ATP channels. These experiments used the fluorescence indicator indo 1 to measure [Ca²⁺]_i and [Ca²⁺]_m in isolated, perfused hearts with parallel measurement of hemodynamics. The role of mitochondrial K_ATP was elucidated using the mitochondrial K_ATP channel opener diazoxide and the mitochondrial K_ATP channel blocker 5-hydroxydecanoic acid (5-HD).

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Heart Perfusion and Measurements of Function

Fluorescence Measurements

Fluorescence instrumentation. Fluorescence measurements were performed as previously described in detail (3, 4) using a modified spectrofluorometer (model SLM8100, SLM Instruments; Rochester, NY). Excitation light from a 450-W xenon arc lamp (Ushio) was filtered through a 350-nm interference filter and focused onto the ingoing leg of a quartz bifurcated fiber bundle (Dolan-Jenner Industries; Lawrence, MA). The common leg of this 1.57-mm diameter fiber bundle was girdled against the epicardial surface of the left ventricle, avoiding visible vessels. A shutter in front of the excitation light was opened for only seconds at a time during data acquisition to prevent bleaching of the indo 1 fluorescence. The fluorescence signal was transferred via the outgoing leg of the bundle and separated by 385- and 456-nm interference filters before detection by photomultiplier tubes.

Indo 1 acetoxymethyl ester loading. After a 15-min equilibration period, baseline background fluorescence [primarily NADH (3, 4)] was measured and subtracted from all subsequent fluorescence measurements. Hearts were then loaded for 25 min by retrograde perfusion with Krebs-Henseleit buffer containing indo 1, acetoxymethyl ester (indo 1-AM; 6 µM, dissolved in dimethyl sulfoxide and Pluronic F-127, 20% wt/vol, Molecular Probes; Johnston City, OR) and fetal bovine serum (1%). Probenecid (0.1 mM, Sigma; St. Louis, MO) was added to all perfusates to slow the extrusion of indo 1 from the myocytes (1). Residual indo 1-AM was washed out by perfusing with standard buffer for 25 min. An experiment was discarded if the fluorescent intensity at either 385 or 456 nm was less than twice the background fluorescence during any phase of the experimental protocol.

Experimental Protocols

Mn²⁺ quenching of cytosolic fluorescence. To determine mitochondrial fluorescence, cytosolic fluorescence was quenched by adding MnCl₂ at a final concentration of 17.5 µM to the perfusate 10 min before the 25-min ischemic period. Adequate quenching was verified by the loss of calcium transients after manganese loading. The addition of MnCl₂ did not alter cardiac function (heart rate or developed pressure).

Calculations

Calculation of [Ca²⁺]_i. [Ca²⁺]_i was calculated using the standard equation for fluorescent calcium indicators (15)
where R is the ratio of fluorescence at 385 nm and 456 nm. R_min and R_max are the fluorescence ratios at zero and saturating [Ca²⁺]_i, respectively, as determined in a previous study (3). A dissociation constant (K_d) of 594 nM was used to calculate [Ca²⁺]_i (2, 34). S₄₅₆ is the ratio of fluorescence intensities during saturating and zero [Ca²⁺]_i at 456-nm emission wavelength (2).

Calculation of [Ca²⁺]_m. [Ca²⁺]_m was calculated using the same equation employed to calculate total [Ca²⁺]_i from the fluorescent intensity after manganese quenching. On the basis of prior studies (28, 29, 35), the calibration parameters were assumed to be the same in the cytosolic and mitochondrial compartments.

NADH autofluorescence. NADH is known to increase during ischemia, thus increasing the background fluorescence signal. Although it was not possible to simultaneously measure NADH and indo 1 fluorescence in each individual heart, NADH fluorescence was measured in parallel experiments in all groups. The change in NADH fluorescence was similar between groups, increasing minimally at 385 nm but doubling at 456 nm and remaining stable during the entire ischemic period. The NADH signal rapidly returned to baseline levels on reperfusion. Thus, whereas [Ca²⁺] might be underestimated due to these changes in NADH after the induction of ischemia, the relative changes should be similar. These experiments also demonstrated that high concentrations of NADH suppressed both excitation and emission light intensities, thus tending to reduce indo 1 fluorescence during ischemia.

Indo 1 leak during ischemia. Loss of membrane integrity could result in the loss of indo 1 during ischemia and reperfusion, yielding both lower intracellular signal intensity as well as providing a signal related to the high extracellular [Ca²⁺] in the perfusate. We measured the amount of indo 1 loss during the initial period of reperfusion and found similar degrees of leakage between groups.

Manganese quenching of indo 1. Because the manganese quench was applied before the onset of ischemia, two issues relating to this procedure arise. First, the extracellular manganese concentration of 17.5 µM is insufficient to fully quench the indo 1 signal at the high (1.2 mM) calcium concentration in the perfusate. In vitro experiments demonstrated that only ~50% of the indo 1 signal was quenched using this concentration of manganese. Second, there is likely leakage of manganese into the mitochondria during ischemia (21, 22), resulting in some quenching of the mitochondrial indo 1 signal. In general, this would result in a loss of the signal-to-noise ratio but no change in the calculated [Ca²⁺]_m, because the fluorescence ratio is used to calculate the concentration. However, differential quenching at different leak amounts could bias the data. Therefore, in vitro experiments were performed at varying concentrations of manganese at a pathophysiological but representative [Ca²⁺] (500 nM). These experiments demonstrated there was no dose effect of manganese on the fluorescent ratio and, hence, calculated [Ca²⁺] (Fig. 1).

View larger version (14K): [in this window] [in a new window]   Fig. 1.   Fluorescence intensity at maximal intensity wavelengths (400 and 470 nm) of indo 1 (5 µM) in a solution of 500 nM [Ca2+] with varying concentrations of MnCl2. Although increasing the manganese concentration lowered the signal intensity at both wavelengths, there was no significant effect on the ratio of intensities (R400/470) and, therefore, on calculated [Ca2+] (inset).

Statistical Analysis

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Hemodynamics

View larger version (29K): [in this window] [in a new window]   Fig. 2.   A: left ventricular developed pressure (LVDP) expressed as a percentage of baseline values. Baseline LVDP was measured after 25 min of washout after 25 min of indo 1 loading. After 30 min of postischemic reperfusion, the recovery of LVDP in the diazoxide (100 µM)-treated and preconditioned (PC) hearts was significantly higher than in the control group. Although 5-hydroxydecanoic acid (5-HD; 100 µM) abolished the effect of diazoxide on LVDP, it only partially inhibited the effects of PC. *P < 0.05 vs. control. B: left ventricular end-diastolic pressure (LVEDP) after 30 min of reperfusion. Reperfusion LVEDP was elevated in control hearts, whereas both diazoxide and PC significantly reduced the reperfusion LVEDP. The effects of diazoxide and PC on LVEDP were completely abolished by 5-HD (100 µM). *P < 0.05 vs. control.

[Ca²⁺]_i

View larger version (25K): [in this window] [in a new window]   Fig. 3.   Intracellular Ca2+ concentration ([Ca2+]i) expressed as a fraction of preischemic baseline value. A: [Ca2+]i in control hearts increased steadily during 25 min of ischemia. The increase in [Ca2+]i was similar between groups except that hearts treated with diazoxide (100 µM) had higher [Ca2+]i after 25 min of ischemia. [Ca2+]i in control hearts increased further during the first 20 min of reperfusion and then fell, but did not return to preischemic levels after 30 min of reperfusion. PC abolished the increase in [Ca2+]i during reperfusion, resulting in significantly lower [Ca2+]i. In contrast, diazoxide increased [Ca2+]i during reperfusion. *P < 0.05 vs. control. B: 5-HD (100 µM) reversed the effects of diazoxide and PC on [Ca2+]i.

[Ca²⁺]_m

View larger version (25K): [in this window] [in a new window]   Fig. 4.   Mitochondrial Ca2+ concentration ([Ca2+]m) expressed as a fraction of preischemic baseline value. A: increase in [Ca2+]m was significantly lower in the diazoxide (100 µM)-treated and PC hearts than in control hearts during 25 min of ischemia and 30 min of reperfusion. *P < 0.05 vs. control. B: 5-HD (100 µM) abolished the effects of both diazoxide and PC on limiting [Ca2+]m overload during ischemia and in the early reperfusion period. However, control hearts still had higher [Ca2+]m in the last 15 min of reperfusion than the 5-HD-treated hearts. *P < 0.05 vs. control.

Relationship of [Ca²⁺] to Functional Recovery

View larger version (15K): [in this window] [in a new window]   Fig. 5.   Relationship between LVDP and [Ca2+]m after 30 min of reperfusion in the 5 groups. Each point is from a different heart. Least squares fits to the points are shown by the solid and dashed lines.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Previous studies using isolated rat hearts have shown that the cardioprotective effect of PC is due in part to opening of mitochondrial K_ATP channels and that diazoxide, a mitochondrial K_ATP channel opener, mimics PC (11, 12, 26). However, the mechanism of this effect is unknown. This study shows that opening mitochondrial K_ATP channels by either PC or a K_ATP channel opener can reduce mitochondrial calcium overload during ischemia and reperfusion in isolated rat hearts. Although the reduction in [Ca²⁺]_m was associated with greater recovery of LVDP, the protective effect was independent of total [Ca²⁺]_i. These data are consistent with the hypothesis that opening mitochondrial K_ATP channels, either pharmacologically or by PC, limits mitochondrial Ca²⁺ overload during ischemia and reperfusion.

[Ca²⁺]_m Overload and Ischemia-Reperfusion Injury

Prior studies have indicated that reperfusion/reoxygenation injury correlates well with [Ca²⁺]_m uptake, although there is no correlation with [Ca²⁺]_i (27, 28). In Miyamae et al. (27), [Ca²⁺]_m was measured by Mn²⁺ quenching 20 min into reperfusion with and without exposure to ruthenium red to block entry of Ca²⁺ into the mitochondria. This study found that functional recovery was greater in hearts exposed to ruthenium red, was inversely correlated with [Ca²⁺]_m, but was independent of [Ca²⁺]_c (which was unaffected by blockade of mitochondrial calcium entry). These findings were consistent with a critical role of pharmacological limitation of [Ca²⁺]_m during reperfusion after ischemia, but did not address whether endogenous limitation of [Ca²⁺]_m occurred during ischemia.

In parallel with pharmacological inhibition of mitochondrial Ca²⁺ entry with ruthenium red (27), the current findings indicate that both PC and K_ATP channel opening with diazoxide limit [Ca²⁺]_m during ischemia as well as reperfusion in parallel with greater functional recovery. As in the prior study (27), the degree of functional recovery was linearly and inversely related to [Ca²⁺]_m below a threshold level (Fig. 5). This may represent a threshold value of [Ca²⁺]_m above which mitochondria cannot function or, possibly, varying proportions of [Ca²⁺]_m values from populations of mitochondria that have normal or lethal calcium concentrations.

The Role of Mitochondrial K_ATP Channels

Previous studies suggest that PC-induced cardioprotection is mediated via activation of mitochondrial K_ATP channels (12, 25). Recently, Fryer et al. (11) examined the role of mitochondrial K_ATP channels in a rat model of PC and demonstrated that inhibition of the mitochondrial K_ATP channel attenuated cardioprotection induced by PC. The present study provides further evidence that PC-induced cardioprotection is indeed mediated by opening of the mitochondrial K_ATP channel in the isolated rat heart because cardioprotection was abolished by the mitochondrial K_ATP channel-selective blocker 5-HD. More importantly, PC-induced opening of mitochondrial K_ATP channels limited [Ca²⁺]_m overload. These data are consistent with the previous hypothesis that PC-induced opening of the mitochondrial K_ATP channel would reduce [Ca²⁺]_m overload (25). As has been proposed, the mitochondrial membrane potential may play an important role in this process. Recently, Ylitalo et al. (38) showed that the decrease in the mitochondrial membrane potential during prolonged ischemia was more rapid in preconditioned hearts. On the basis of these data, we can speculate that lower [Ca²⁺]_m in the preconditioned hearts resulted, at least in part, from a decrease in the driving force for Ca²⁺ entering mitochondria through the Ca²⁺ uniporter.

Although the current results do not indicate whether lower [Ca²⁺]_m during ischemia and reperfusion was directly responsible for the cardioprotective effect of the mitochondrial K_ATP channel opener and PC, previous studies (27) with Ca²⁺-uniporter blockers suggest this is likely an important component of cardioprotection. By assessing mitochondrial Ca²⁺ and the functional recovery in isolated rat hearts, this study provides a link between ischemia-reperfusion injury, [Ca²⁺]_m overload, and mitochondrial K_ATP channel activation.

Mitochondrial Versus Cellular Calcium Overload

The role of lower [Ca²⁺]_i in the PC hearts and the mechanism of this observation is only speculative. It is unclear whether the limitation of [Ca²⁺]_i was a primary effect of PC or, more likely, a secondary effect of improved mitochondrial function resulting in greater ATP production and, hence, sarcolemmal Ca-ATPase function (10). Alternatively, other mechanisms of PC, such as preservation of sarcoplasmic reticulum (SR) function, may play a role in regulating [Ca²⁺]_i. Investigators (31, 36) have reported that SR Ca²⁺ release and Ca²⁺ uptake activity was relatively preserved in PC hearts after reperfusion, an effect that may result in greater SR accumulation of calcium and lower cytosolic calcium. Another factor that may affect intracellular calcium in PC hearts is the rapid reduction in intracellular sodium observed during reperfusion (33), an event that would alter the kinetics of Na⁺/Ca²⁺ exchange across the sarcolemma and limit [Ca²⁺]_i. In contrast, the pharmacological opening of mitochondrial K_ATP channels with diazoxide would not be expected to produce these ancillary effects of PC and, by limiting mitochondrial Ca²⁺ accumulation, result in higher [Ca²⁺]_i.

These data suggest that elevation of the intracellular level of [Ca²⁺]_i may not be necessarily pathological when the mitochondrial level of [Ca²⁺]_m is limited. This may occur because a critical element in determining cellular recovery and the return of ion homeostasis on reperfusion is the ability of the cell to synthesize ATP and maintain the free energy of hydrolysis (G) above a critical level.

Limitations

The present study has shown that PC and diazoxide each protect the myocardium against ischemia-reperfusion injury and enhance the recovery of postischemic contractile function. The mechanism underlying this cardioprotective effect of diazoxide and PC may be attributed to limiting mitochondrial, but not intracellular, Ca²⁺ overload during ischemia and reperfusion by opening mitochondrial K_ATP channels.