INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

ATP-REGULATED POTASSIUM (K_ATP) channels were first described by Noma (27) in cardiac myocytes (7) and mediate the protective effects of ischemic preconditioning (11). More recently, the mitochondrial K_ATP channel has been suggested to be critically important as an effector of endogenous cardioprotective signal transduction (19). Pancreatic -cells also contain K_ATP channels that regulate insulin release (7) and close in response to glucose (2). The regulation of cardiac mitochondrial K_ATP channels by glucose has not been evaluated. Our recent findings demonstrate that hyperglycemia produced by administration of glucose (15) or diabetes (16) abolishes the reductions of myocardial infarct size produced by ischemic preconditioning. This may be secondary to glucose impairing activation of mitochondrial K_ATP channels. Therefore, we tested the hypothesis that diabetes or hyperglycemia prevents decreases in infarct size produced by the mitochondrial K_ATP channel agonist diazoxide (19).

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

All experimental procedures and protocols used in this investigation were reviewed and approved by the Animal Care and Use Committee of the Medical College of Wisconsin. All conformed to the "Guiding Principles in the Care and Use of Animals" of the American Physiological Society and the Guide for the Care and Use of Laboratory Animals [(7th ed.) Washington, DC: Nat. Acad. Press, 1996].

General preparation. The surgical instrumentation has been described previously (15, 16). Briefly, mongrel dogs (weight 23 ± 2 kg; means ± SE) were anesthetized with barbital sodium (200 mg/kg) and pentobarbital sodium (15 mg/kg), intubated, and ventilated with oxygen-enriched air (FI_O2 = 0.25) with the use of positive pressure ventilation. Tidal volume and respiratory rate were adjusted during each experiment to maintain arterial blood gas tensions and acid-base status near conscious values. A double-pressure transducer-tipped catheter was inserted into the aorta and left ventricle (LV) via the left carotid artery to measure arterial and LV pressures, respectively, and the maximum rate of increase of LV pressure (dP/dt_max). The right femoral vein and artery were catheterized for fluid administration and withdrawal of reference arterial blood samples, respectively. A thoracotomy was performed in the left fifth intercostal space, the lung was retracted, and the heart was suspended in a pericardial cradle. A 1.5- to 2-cm segment of the proximal left anterior descending (LAD) coronary artery distal to the first diagonal branch was isolated, and a silk ligature was placed around the vessel for production of coronary artery occlusion and reperfusion. Myocardial infarct size was determined with triphenyltetrazolium chloride (TTC) staining at the completion of each experiment (16). TTC is reduced by dehydrogenase enzymes present in viable (noninfarcted) myocardium to a formazan precipitate, turning viable tissue a brick-red color. Infarcted tissue, depleted of dehydrogenase enzymes during reperfusion, retains a pale gray appearance (32). A catheter was placed in the left atrium for injection of radioactive microspheres that were used to measure regional myocardial perfusion in the ischemic (LAD) zone at selected intervals (16). All hemodynamic data were monitored on a polygraph and digitized with a computer interfaced with an analog-to-digital converter.

Experimental protocol. Ninety minutes after instrumentation was completed, baseline systemic hemodynamics were recorded. All dogs were subjected to a 60-min LAD occlusion followed by 3 h of reperfusion. In two experimental groups (Fig. 1), the dogs received drug vehicle (saline) or intravenous diazoxide (2.5 mg/kg over 15 min). The effects of diabetes on myocardial infarct size were investigated in two additional groups of dogs. Diabetic dogs received intravenous alloxan (40 mg/kg) and streptozotocin (25 mg/kg) and were studied with and without diazoxide 3 wk after induction of diabetes as previously described (16). Only dogs with sustained hyperglycemia (blood glucose > 200 mg/dl) without ketosis or renal insufficiency (blood urea nitrogen < 40 mg/dl) were studied. The effects of acute hyperglycemia on infarct size were evaluated in two groups of dogs by infusing 15% dextrose in water to increase blood glucose to 300 mg/dl (moderate hyperglycemia) in the presence or absence of pretreatment with diazoxide. To address the dose-dependent effects of diazoxide and glucose on myocardial infarct size, three final groups of dogs received high-dose diazoxide (5.0 mg/kg) in the absence or presence of moderate or profound (600 mg/dl) hyperglycemia. Dogs that developed intractable ventricular fibrillation and those with subendocardial coronary collateral blood flow 0.15 ml · min¹ · g¹ were excluded from data analysis (11, 25).

View larger version (20K): [in this window] [in a new window]   Fig. 1.   Schematic diagram illustrating the experimental protocol.

Statistical analysis. Statistical analyses of data within and among groups were performed with a two-way ANOVA for repeated measures followed by application of Student-Newman-Keul's test. Myocardial infarct size was evaluated with one-way ANOVA. The relationship between blood glucose or transmural collateral blood flow and infarct size was evaluated with multiple regression analysis. Changes within and among groups were considered statistically significant when P < 0.05. Data are expressed as means ± SE.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Seventy-eight dogs were instrumented to obtain sixty-two successful experiments. Four dogs were excluded because of intractable ventricular fibrillation [1 control, 2 diazoxide (2.5 mg/kg), and 1 hyperglycemia + diazoxide (2.5 mg/kg) dog]. Twelve dogs were excluded because subendocardial blood flow exceeded 0.15 ml · min¹ · g¹ [2 control, 2 diabetic, 2 moderate hyperglycemia, 2 diazoxide (2.5 mg/kg), 2 diabetes + diazoxide (2.5 mg/kg), and 2 moderate hyperglycemia + diazoxide (2.5 mg/kg) dogs]. Ischemia and reperfusion had no effect on blood glucose concentrations during control experiments (n = 3, 75 ± 3, 69 ± 3, and 76 ± 6 mg/dl at baseline, during coronary occlusion, and after 3 h of reperfusion, respectively). Blood glucose concentrations were similar in diabetic (229 ± 31 mg/dl) and moderately hyperglycemic (310 ± 10 mg/dl) dogs and in the presence of diazoxide [diabetes + diazoxide: 239 ± 18 mg/dl; moderate hyperglycemia + diazoxide (2.5 mg/kg): 329 ± 17 mg/dl; moderate hyperglycemia + diaxozide (5.0 mg/kg): 397 ± 16 mg/dl]. Diazoxide alone had no effect on blood glucose concentrations (n = 6, 68 ± 2, 72 ± 2, and 72 ± 4 mg/dl at baseline, during coronary artery occlusion, and after 3 h of reperfusion, respectively). Blood glucose concentrations were more variable in diabetic animals (range: 152-338 mg/dl). During profound hyperglycemia, blood glucose concentrations were 574 ± 23 mg/dl.

Hemodynamics. There were no differences in hemodynamics (Table 1) among groups at baseline, preocclusion, and during coronary artery occlusion or reperfusion. Diazoxide decreased mean arterial pressure in the absence or presence of hyperglycemia. Coronary artery occlusion and reperfusion increased LV end-diastolic pressure and decreased dP/dt_max in each group. There were no differences in arterial blood gases (Table 2) among groups.

Myocardial infarct size. The LV area at risk (AAR) for infarction was similar between groups [control, 39 ± 2; diabetic, 44 ± 2; moderate hyperglycemia, 38 ± 2; diazoxide (2.5 mg/kg), 44 ± 2; diabetes + diazoxide, 45 ± 2; moderate hyperglycemia + diazoxide (2.5 mg/kg), 43 ± 1; diazoxide (5.0 mg/kg), 40 ± 4; moderate hyperglycemia + diazoxide (5.0 mg/kg), 42 ± 2; profound hyperglycemia + diazoxide (5.0 mg/kg), 42 ± 3%]. Myocardial infarct size expressed as a percentage of the AAR (Fig. 2) was 25 ± 1% in control experiments (n = 8) and was significantly (P < 0.05) reduced by diazoxide (n = 9) to 10 ± 1%. Infarct size was similar in diabetic (n = 8, 28 ± 3%) and hyperglycemic dogs (n = 8, 25 ± 1%) compared with controls. Diazoxide did not protect against infarction in diabetic (n = 9, 28 ± 5%) or moderately hyperglycemic (n = 8, 23 ± 2%) dogs. High-dose diazoxide (n = 4) caused similar reductions in infarct size (11 ± 2%) compared with lower doses; however, moderate hyperglycemia (n = 4) did not abolish the protection of high-dose diazoxide (15 ± 3%). In contrast, profound hyperglycemia (n = 4) blocked reductions in infarct size produced by high-dose diazoxide (26 ± 3%).

View larger version (16K): [in this window] [in a new window]   Fig. 2.   Myocardial infarct size expressed as a percentage of left ventricular (LV) area at risk (AAR) for infarction in control (Con), diabetic (DM), and hyperglycemic (H300; blood glucose concentration 300 mg/dl) dogs in the absence and presence of 2.5 mg/kg diazoxide (Diaz2.5). *Significantly (P < 0.05) different from Con; significantly (P < 0.05) different from DM; significantly (P < 0.05) different from H300; ¶Significantly (P < 0.05) different from DM + Diaz2.5; §significantly (P < 0.05) different from H300 + Diaz2.5.

Myocardial perfusion. Myocardial blood flow in the ischemic region is summarized in Table 3. There were no significant differences in baseline or coronary collateral blood flow among groups. Infarct size was inversely related to coronary collateral blood flow in control but not diazoxide-treated (Fig. 3), diabetic (Fig. 4), or hyperglycemic dogs (data not shown). In contrast, blood glucose was the most important determinant of myocardial infarct size in diabetic (Fig. 5) and hyperglycemic dogs (Fig. 6).

View larger version (20K): [in this window] [in a new window]   Fig. 3.   Relationships between myocardial infarct size expressed as a percentage of the LV AAR for infarction and transmural collateral blood flow in control dogs and those treated with 2.5 mg/kg diazoxide. Myocardial infarct size was inversely related to collateral blood flow in control (P < 0.05) but not diazoxide-treated dogs. Diazoxide significantly (P < 0.05) shifted the regression line downward compared with control experiments, indicating cardioprotection.

View larger version (15K): [in this window] [in a new window]   Fig. 4.   Relationship between myocardial infarct size expressed as a percentage of the LV AAR for infarction and transmural collateral blood flow in diabetic dogs in the presence or absence of 2.5 mg/kg diazoxide. Infarct size was unrelated to collateral blood flow in diabetic dogs.

View larger version (22K): [in this window] [in a new window]   Fig. 5.   Relationship between myocardial infarct size expressed as a percentage of the LV AAR for infarction and blood glucose concentration in diabetic dogs in the presence and absence of 2.5 mg/kg diazoxide. Thick solid line, regression relationship previously determined in diabetic dogs (y = 0.07x + 6.99; r = 0.69) (16). A similar, although not statistically significant, relationship was observed in diabetic dogs during the current investigation in the absence (thin solid line; P = 0.07) or presence (dashed line; P = 0.29) of diazoxide. When data from all diabetic dogs were pooled, a statistically significant relationship between infarct size and blood glucose was demonstrated (P < 0.02).

View larger version (16K): [in this window] [in a new window]   Fig. 6.   Myocardial infarct size expressed as a percentage of LV AAR for infarction in dogs receiving 5.0 mg/kg diazoxide (Diaz5.0). Diazoxide significantly (P < 0.05) reduced infarct size compared with control experiments. There was no difference in the extent of infarct reduction by 2.5 or 5.0 mg/kg diazoxide. Profound (H600; blood glucose concentration 600 mg/dl) but not moderate (H300) hyperglycemia blocked the protective effects of high-dose diazoxide.*Significantly (P < 0.05) different from Diaz5.0; significantly (P < 0.05) different from H300 + Diaz5.0.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Diabetes mellitus adversely affects the prognosis of patients with acute myocardial infarction. Despite significant advances in the treatment of coronary artery disease, the mortality rate associated with myocardial infarction is two to three times greater in diabetic patients compared with those without this disorder (1, 21). The degree of metabolic control (18) and not the presence of conventional risk factors appears to be of major importance in determining the outcome of diabetic patients with infarction (22, 24). Moreover, hyperglycemia on hospital admission is an independent risk factor predicting an increase in mortality rate in patients with or without diabetes (4, 14, 22). Thus diabetes and hyperglycemia are clear contributors to adverse outcome in patients at risk for myocardial ischemia, but the mechanisms responsible for this observation are poorly understood.

Ischemic preconditioning is a powerful endogenous mechanism protecting the myocardium against infarction (25). Ischemic preconditioning occurs in patients with coronary artery disease (17, 31) and may be adversely affected by K_ATP channel antagonists, including sulfonylurea hypoglycemic agents. Chronic treatment with sulfonylureas has been shown to abolish the protection afforded by preconditioning in the human myocardium harvested from diabetic but not insulin-treated or normal patients (6). Previous work in dogs also indicates that diabetes (16) or hyperglycemia alone (15) prevents reductions of myocardial infarct size by ischemic preconditioning. In contrast, other studies (20, 30) have suggested that the diabetic rat or rabbit myocardium may be resistant to infarction. However, the degree of hyperglycemia observed during the latter investigations may have been insufficient to result in a deleterious effect. For example, fasting blood glucose concentrations were near normal (131 ± 8 mg/dl) in diabetic rabbits (20), and isolated hearts from diabetic rats were perfused with normoglycemic crystalloid solutions during experimentation in another study (30). A critical role of glucose to modulate myocardial infarct size was recently demonstrated by a linear relationship between blood glucose concentration and infarct size in diabetic or acutely hyperglycemic dogs that persists with or without preconditioning stimuli (16). The present results confirm these findings and demonstrate that hyperglycemia caused by administration of exogenous dextrose or diabetes is an important determinant of myocardial infarct size during activation of mitochondrial K_ATP channels. Indeed, the dose of diazoxide and degree of hyperglycemia are interactive. The protection afforded by low-dose diazoxide is blocked by moderate hyperglycemia, whereas profound degrees of hyperglycemia are required to block high-dose diazoxide. These findings occurred independent of alterations in systemic hemodynamics or coronary collateral blood flow. Similar to our previous findings, myocardial infarct size was not related to coronary collateral blood flow during diabetes or hyperglycemia (15, 16) but was related to the degree of hyperglycemia present. The relationship between blood glucose concentration and myocardial infarct size was similar in diabetic animals receiving diazoxide; however, the correlation between these variables was weaker. The results are also in agreement with recent work demonstrating that diazoxide does not precondition atrial muscle harvested from diabetic patients (10). Taken together, these data suggest that hyperglycemia or diabetes may adversely affect mitochondrial K_ATP channel function and exacerbate myocardial ischemic injury. The data also suggest that activation of mitochondrial K_ATP channels may be possible during diabetes if blood glucose concentrations are maintained near normal.

The results of the current investigation could be interpreted to conflict with clinical studies advocating the infusion of glucose-insulin-potassium (GIK) to patients at risk for myocardial ischemia (8). The findings of these studies should be interpreted within the framework of glycemic control, however. Blood glucose concentrations were not consistently controlled or reported in many previous investigations. Interestingly, one trial designed to evaluate the efficacy of GIK for the treatment of acute myocardial infarction was discontinued for ethical reasons because mortality was increased in the group receiving GIK (13). More importantly, blood glucose concentrations were poorly controlled in patients receiving GIK (500-600 mg/dl) in this study compared with the control group or compared with another study demonstrating a beneficial effect of GIK (peak blood glucose concentration <300 mg/dl) (29). Insulin itself has also been shown to be cardioprotective by a mechanism independent of K_ATP channels (3). Thus the current results suggest that glycemic control may be an important variable that determines the efficacy of GIK to reduce myocardial ischemic injury.

The results of the current investigation should be interpreted within the constraints of several potential limitations. The LV AAR for infarction and coronary collateral blood flow are important determinants of the extent of myocardial infarction. However, there were no differences in these variables among experimental groups that may account for the present findings. The dose of diazoxide determined in preliminary experiments decreased mean arterial pressure before and during LAD occlusion. This reduction in arterial pressure may have contributed to a decline in myocardial oxygen consumption and a subsequent decrease in infarct size. However, hyperglycemia and diabetes attenuated the protective effects of diazoxide despite producing similar hemodynamic effects. A role for sarcolemmal K_ATP channel activation by diazoxide during the current investigation, while unlikely, cannot be completely excluded. Diazoxide is reported to be relatively selective for mitochondrial K_ATP channels (9); however, Birincioglu et al. (5) recently reported that the anti-infarct effects of a higher dose of diazoxide (10 mg/kg) were blocked by a selective sarcolemmal K_ATP channel antagonist. The nature of the interaction between hyperglycemia and mitochondrial K_ATP channels is also unresolved. Whereas extensive experimental evidence supports the contention that K_ATP channels are the distal effectors of ischemic preconditioning (12), recent evidence suggests that opening of mitochondrial K_ATP channels may, instead, trigger a preconditioned state (28). Whether hyperglycemia specifically prevents such triggering or whether hyperglycemia disrupts signaling more distal in the preconditioning pathway remains to be determined. Finally, the extent of infarct reduction by diazoxide may have been overestimated using TTC staining techniques. This is a standard method of determining myocardial infarct size (5, 11, 26, 28, 32), but it has recently been suggested that infarct size may be underestimated after brief periods of reperfusion (23). Decreases in myocardial infarct size (determined by TTC staining) were demonstrated after 3 h of reperfusion in rabbits subjected to late ischemic preconditioning stimuli but were not observed when the myocardium was evaluated with histological techniques after 3 days of reperfusion. Whether this discrepancy occurred because preconditioning merely delayed the progression to infarction or because TTC staining underestimated infarct size was not determined. Nevertheless, this limitation has not been demonstrated in the canine myocardium, and diazoxide has been shown to cause similar decreases in cell death in a model independent of reperfusion duration or use of TTC staining techniques (19).

In conclusion, the present results demonstrate that diabetes and acute hyperglycemia attenuate reductions of myocardial infarct size produced by the selective mitochondrial K_ATP channel agonist diazoxide. These findings suggest that glucose may be a critical modulator of mitochondrial K_ATP channels during myocardial ischemia. The results implicate impaired cardioprotective signal transduction as a likely mechanism for increased cardiovascular risk during diabetes mellitus.