INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

UNDER NORMAL PHYSIOLOGICAL CONDITIONS, fatty acid oxidation is responsible for about 50 ~ 70% of the energy production of the heart, with the remainder arising from glucose metabolism (see Ref. 29 for review). However, in uncontrolled diabetes, glucose utilization is markedly impaired, and almost all of the energy supply in the heart originates from fatty acid oxidation (12, 13, 29, 31). The reason for this switch from glucose to fatty acid metabolism may partly be explained by changes in circulating substrates and insulin levels in the diabetic. Although glucose levels are elevated in uncontrolled diabetes, myocardial glucose metabolism is impaired due to a decrease in glucose uptake into the myocyte by insulin-dependent GLUT-4 transporters. High circulating fatty acid levels that can be seen in the diabetic (23) can also potentially increase fatty acid oxidation rates, with a parallel inhibition of glucose oxidation (26, 31). However, isolated perfused heart studies have suggested that the dramatic alterations in cardiac energy metabolism in the diabetic cannot be explained entirely by alterations in circulating substrate or insulin levels (2, 26). In particular, alterations in the subcellular control of fatty acid oxidation may contribute to the switch from glucose to fatty acid metabolism in the diabetic heart (12, 13). To date, it has not been completely delineated what alterations in the control of fatty acid oxidation occur in the diabetic.

Carnitine palmitoyltransferase (CPT) 1, a key enzyme involved in fatty acid oxidation, is the rate-limiting enzyme involved in mitochondrial fatty acid uptake (29, 33). Malonyl-CoA, which is produced by acetyl-CoA carboxylase (ACC), is a potent inhibitor of CPT 1 and acts at a site distinct from the catalytic site of CPT 1. Although the sensitivity of cardiac CPT 1 to malonyl-CoA inhibition does not change in the diabetic (3), a decrease in malonyl-CoA content has been seen in streptozotocin-induced diabetic swine hearts (7), suggesting less malonyl-CoA inhibition of CPT 1 and therefore greater rates of fatty acid oxidation. This raises the possibility that decreased production and/or increased degradation of malonyl-CoA reduces malonyl-CoA levels and contributes to the high fatty acid oxidation rates seen in the diabetic heart.

Studies from our laboratory (5, 10, 11, 21) and others (1) established that ACC is a very important determinant of malonyl-CoA levels and fatty acid oxidation rates in the heart. A key kinase responsible for the control of ACC activity is AMP-activated protein kinase (AMPK). As a result, AMPK is an important regulator of fatty acid oxidation in the heart, since it phosphorylates and inactivates ACC, resulting in a decrease in malonyl-CoA production and an increase in fatty acid oxidation rates (5, 10).

Although the control of malonyl-CoA synthesis has been extensively studied in the heart, fewer studies have addressed how malonyl-CoA is degraded. Our recent studies have shown that the heart contains an active malonyl-CoA decarboxylase (MCD) that decarboxylates malonyl-CoA back to acetyl-CoA (5). In newborn hearts, the increase in fatty acid oxidation that occurs after birth is associated with an increase in MCD activity (5). What effect diabetes has on MCD activity and expression has not been determined.

The purpose of the present study was to determine the importance of diabetes-induced extracellular changes in energy substrate levels and insulin levels on cardiac energy substrate preference versus intracellular changes in the control of fatty acid metabolism. This was achieved by measuring glucose and fatty acid oxidation in isolated working hearts from streptozotocin diabetic rats perfused in the presence or absence of insulin, under either diabetic conditions (high fatty acids/high glucose) or control conditions (low fatty acids/low glucose). We demonstrate that changes in subcellular control of fatty acid oxidation, possibly at the level of MCD, are primarily responsible for the excessive reliance on fatty acid oxidation in the diabetics, as opposed to alterations in circulating insulin and substrate levels.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Animals. Male Sprague-Dawley rats (230-250 g) were anesthetized with enflurane and then were given a single tail vein injection of 55 mg/kg streptozotocin (Sigma Chemicals, St. Louis, MO). Streptozotocin was dissolved in 50 mM citrate, pH 4.5. Control animals were injected with the vehicle alone. Animals were allowed to recover and were administered food and water ad libitum for a 6-wk period. Only animals demonstrating urine glucose >2% within 48 h of streptozotocin injection were used in the diabetic group.

Heart perfusions. Hearts from pentobarbital sodium-anesthetized 6-wk control and diabetic rats were excised, the aorta was cannulated, and a retrograde perfusion with Krebs-Henseleit solution (pH 7.4, gassed with 95% O₂-5% CO₂) was initiated, as described previously (17). During this initial perfusion, the hearts were trimmed of excess tissue, the pulmonary artery was cut, and the opening to the left atrium was cannulated. After a 10-min Langendorff washout period, hearts were switched to the working mode and were perfused at a left atrial preload of 11.5 mmHg and an aortic afterload of 80 mmHg. Spontaneously beating hearts were used throughout the study, with left ventricular peak systolic pressure measured with a Spectramed P23 XI pressure transducer in the aortic afterload line and recorded on a Gould RS-3600 physiograph. Cardiac output and aortic flow were measured with Transonic ultrasonic flow probes present in the atrial preload line and the aortic outflow lines, respectively. Coronary flow was calculated as the difference between cardiac output and aortic flow. Isolated hearts from control or diabetic rats were perfused with Krebs-Henseleit solution containing 3% BSA that contained ~0.05 mM free fatty acids, 2.5 mM free Ca²⁺, and either 5 mM [U-¹⁴C]glucose and 0.4 mM [9,10-³H]palmitate (low-fat/low-glucose buffer) or 20 mM [U-¹⁴C]glucose and 1.2 mM [9,10-³H]palmitate (high-fat/high-glucose buffer) in the presence or absence of 100 µU/ml insulin. The radiolabeled palmitate (0.4 or 1.2 mM) was prebound to the albumin (Boehinger Mannheim Fraction V). Fatty acid content of this BSA was premeasured and found to contain 0.05 mM fatty acids/3% BSA. The concentration of glucose and palmitate used in high-fat/high-glucose buffer was based on our previous study that an injection of 55 mg/kg streptozotocin results in an increase in serum glucose to 23.5 ± 5.6 mM and an increase in serum free fatty acids to 1.23 ± 0.14 mM (16).

Measurement of glucose oxidation. Steady-state glucose oxidation was measured by quantitatively collecting ¹⁴CO₂ produced by the heart from [U-¹⁴C]glucose in the buffer, as described previously (19). The ¹⁴CO₂ produced as a gas was determined by bubbling gas in the closed perfusion system through a 1 mol/l methylbenzethonium hydroxide ¹⁴CO₂ trap. This allowed for the quantitative collection of all ¹⁴CO₂ released as a gas. The hyamine hydroxide solution was sampled at 10-min intervals during the 40-min period in which glucose oxidation was measured. At the same time, duplicate perfusate samples were collected and stored under mineral oil to prevent the liberation of ¹⁴CO₂, and the ¹⁴CO₂ was subsequently extracted from the [¹⁴CO₂]bicarbonate, as described previously (19). This involved injecting perfusate samples in closed metabolic reaction flasks containing 9 N H₂SO₄ and gently shaking for 1 h. The ¹⁴CO₂ released from the perfusion buffer was trapped in center wells filled with 1 mol/l methylbenzethonium hydroxide. The center wells were subsequently removed and counted in ACS scintillant by using standard -scintillation counting procedures. Glucose oxidation rates were expressed as nanomoles of glucose per minute per gram dry heart weight.

Measurement of palmitate oxidation. Palmitate oxidation was determined by measuring the ³H₂O content in the perfusate samples as previously described (25). ³H₂O was separated from [9,10-³H]palmitate by treating 0.5-ml buffer samples with 1.88 ml of a mixture of chloroform-methanol and KCl-HCl (1:2 vol/vol) and then adding 0.625 ml of chloroform and 0.625 ml of a 2 M KCl-HCl solution. The aqueous phase was collected using a Pasteur pipette and was subsequently treated with a mixture of chloroform, methanol, and KCl-HCl (1:1:0.9). Two 0.5-ml samples of aqueous phase were counted for each perfusate sample for total ³H₂O determination, taking into account the dilution factor. This method results in >99.7% extraction and separation of ³H₂O from the [9,10-³H]palmitate. Palmitate oxidation rates were expressed as nanomoles of palmitate oxidized per minute per gram dry heart weight.

Tissue workup. The frozen ventricular tissue was weighed and powdered in a mortar and pestle cooled to the temperature of liquid N₂, and the tissue was stored in liquid N₂ for subsequent biochemical analysis. CoA esters were extracted from the powdered tissue using 6% perchloric acid, as described previously (24). The CoA esters were separated and quantified using a previously described high-performance liquid chromatography procedure (8).

Extraction of AMPK and ACC. Approximately 200 mg of frozen tissue were homogenized with a buffer containing Tris · HCl (0.05 M/l, pH 7.5 at 4°C), mannitol (0.25 M/l), NaF (50 mM), sodium pyrophosphate (5 mM), EDTA (1 mM), EGTA (1 mM), dithiothreitol (1 mM), and the following protease inhibitors: phenylmethylsulfonyl fluoride (1 mM), soybean trypsin inhibitor (4 µg/ml), and benzamidine (1 mM). Samples were centrifuged at 14,000 g for 20 min at 4°C. The supernatant was brought to 2.5% polyethylene glycol (PEG) with 25% (wt/vol) PEG 6000 and was agitated for 10 min at 4°C. Samples were then spun at 10,000 g for 10 min at 4°C. The supernatant was the made up to 6% PEG 6000 using PEG 6000 stock described above and was stirred once again for 10 min at 4°C. This fraction was spun at 10,000 g for 10 min, and the precipitate was washed with homogenization buffer containing 6% PEG 6000. This was followed by a final centrifugation at 10,000 g, after which the protein concentration in the supernatant was measured using a Sigma bicinchoninic acid protein kit.

ACC assay. ACC activity in the PEG 6000 fractions was measured using the CO₂ fixation method (32). Briefly, 5 µl of the PEG fraction, containing 20 µg of total protein, were added to a reaction mixture (final volume, 165 µl) containing Tris acetate (60.6 mM), BSA (1 mg/ml), 2-mercaptoethanol (1.32 µmol/l), ATP (2.21 mM), acetyl-CoA (1.06 mM), magnesium acetate (5.0 mM), and NaHCO₃ (18.08 mM). Samples were incubated at 37°C for 10 min, and the reaction was stopped by adding 25 µl of 10% perchloric acid. Samples were then spun for 20 min at 3,500 rpm, and 160 µl of supernatant were placed in minivials and dried in a fume hood overnight. H₂O (100 µl), followed by scintillant, was added to the vials, and the vials were counted. ACC activity was expressed as the amount of malonyl-CoA produce per minute per milligram protein.

AMPK assay. AMPK activity was measured by following the incorporation of ³²P into a synthetic peptide (10). Briefly, 2 µl of the PEG fraction were added to a reaction mixture (final volume, 25 µl) composed of HEPES-NaOH (40 mM), NaCl (80 mM), glycerol (8%, wt/vol), EDTA (0.8 mM), AMARAASAAALARRR (AMARA) peptide (200 µmol/l), dithiothreitol (0.8 mM), [-³²P]ATP (200 µmol/l), MgCl₂ (5 mM), and 0.18% Triton X-100. Samples were also incubated in the presence or absence of 200 µmol/l AMP. This mixture was incubated for 3 min at 30°C. From this incubation mixture, 15 µl were spotted on 1-cm² phosphocellulose paper. The paper was then washed four times for 10 min each with 150 mM phosphoric acid, followed by a 5-min acetone wash. Papers were then dried and counted for radioactivity. AMPK activity was expressed as picomoles ³²P incorporated in the AMARA peptide per minute per milligram protein.

MCD assay. MCD activity was measured by detecting acetyl-CoA, the product of the MCD reaction. Acetyl-CoA derived from MCD was incubated in the presence of [¹⁴C]oxaloacetate and citrate synthase (0.73 µU/µl) to form citrate. The [¹⁴C]oxaloacetate was initially produced by a 20-min transamination reaction performed at room temperature using L-[U-¹⁴C]aspartate (2.5 µCi/ml) and 2-oxoglutarate (2 mM). One of the advantages of this assay was that enzyme activity reflected the rate of enzyme activity in vivo. To initiate the MCD assay, heart homogenates were incubated in a 210-µl reaction mixture (0.1 M Tris, pH 8; 0.5 mM dithiothreitol; 1 mM malonyl-CoA) for 10 min at 37°C, in the presence or absence of NaF (50 mM) and NaPPi (5 mM). The reaction was stopped by the addition of 40 µl of perchloric acid (0.5 mM), neutralized with 10 µl of 2.2 M KHCO₃ (pH 10), and centrifuged at 10,000 g for 5 min to remove precipitated proteins. The incubation of the heart sample with malonyl-CoA allowed for the conversion of malonyl-CoA to acetyl-CoA, which was then combined with [¹⁴C]oxaloacetate (0.17 µCi/ml) to produce [¹⁴C]citrate. All reactions were carried out in the presence of N-ethylmaleimide, which removes excess CoA remaining in the latter stages of the reaction so that the citrate present could not generate non-malonyl-CoA decarboxylase-derived acetyl-CoA. Unreacted [¹⁴C]oxaloacetate was removed from the reaction mixture by the addition of sodium glutamate (6.8 mM) and aspartate aminotransferase (0.533 µU/µl), followed by a 20-min incubation at room temperature. This allows for transamination of unreacted [¹⁴C]oxaloacetate back to [¹⁴C]aspartate. The resulting solution was then stirred in a 1:2 suspension of Dowex fraction-removed [¹⁴C]aspartate, whereas the supernatant contained [¹⁴C]citrate. The supernatant fraction was then counted for ¹⁴C present in the form of [¹⁴C]citrate. The amount of acetyl-CoA produced by MCD was then quantified by comparison with acetyl-CoA standard curves that had been subjected to the identical assay conditions described above. A standard acetyl-CoA concentration curve was run with each experiment. These curves were always found to be linear (r = 0.99, data not shown).

Western blot analysis. Samples were subjected to SDS-PAGE and were transferred to nitrocellulose as described previously (5). Membranes were immunoblotted with rat anti-MCD and were visualized using the Amersham Enhanced Chemiluminescence Western blotting detection system.

Statistical analysis. Data are expressed as the means ± SE. ANOVA followed by the Newman-Keuls test was used to determine statistical significance. Comparison between two groups was made using the Student's t-test. Significance was set at P < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

General features of control and streptozotocin-induced diabetic rats. Table 1 shows body weight and plasma glucose and fatty acid concentrations determined 6 wk after injection of rats with either vehicle (control) or 55 mg/kg iv streptozotocin (diabetic). As expected, diabetic rats had a significantly lower body weight at the time of death compared with age-matched control rats. Plasma glucose levels of diabetic animals were also significantly higher than control rats, whereas plasma insulin levels were significantly decreased. Serum fatty acid levels were also significantly elevated in diabetic compared with control rats.

Hemodynamic data. To determine the importance of circulating glucose and fatty acids on energy metabolism, hearts from control and diabetic rats were perfused with either normal physiological levels of glucose and fatty acids (5 mM glucose and 0.4 mM palmitate) or levels of glucose and fatty acids that can be seen in the uncontrolled diabetic (20 mM glucose and 1.2 mM palmitate). These perfusions were also performed either in the absence or presence of 100 µU/ml of insulin. Table 2 summarizes the cardiac function seen in the isolated working hearts. No deterioration in mechanical function was seen throughout the 40-min perfusion period in any of the experimental groups (data not shown). Consistent with previous studies (6), heart rate was depressed in diabetic rat hearts, although this did not reach statistical significance in the diabetic group perfused with high fat/high glucose and 100 µU/ml insulin. Peak systolic pressure was not different between any perfusion group, although rate pressure product (RPP) tended to be lower in the diabetic groups due to the lower heart rate observed in these rats. A significant decrease in RPP was observed in the low-fat/low-glucose diabetic group perfused with 100 µU/ml compared with control hearts perfused under similar conditions. No significant difference in cardiac output, aortic flow, or cardiac work was seen between the perfusion groups.

Glucose and palmitate oxidation rates in control and diabetic rats. The effect of altering glucose, fatty acid, and insulin concentration on glucose oxidation rates (A) and palmitate oxidation rates (B) in control and diabetic rats is shown in Fig. 1. Addition of insulin to control hearts perfused with low fat/low glucose resulted in a significant increase in glucose oxidation rates. Insulin increased glucose oxidation in control hearts perfused with low-fat/low-glucose buffer and tended to decrease palmitate oxidation, but this reduction was not statistically significant.

View larger version (19K): [in this window] [in a new window]   Fig. 1.   Rates of glucose oxidation (A) and palmitate (B) oxidation in isolated working hearts from control and diabetic rats. Values represent means ± SE of 10 hearts in each group. Hearts were aerobically perfused with either 5 mM glucose and 0.4 mM palmitate (low-fat/low-glucose buffer) or 20 mM glucose and 1.2 mM palmitate (high-fat/high-glucose buffer). *Significantly different from control hearts perfused with low-fat/low-glucose buffer in the absence of insulin. #Significantly different from diabetic hearts perfused with low-fat/low-glucose buffer in the absence of insulin.

Tricarboxylic acid cycle activity. The contribution of glucose and palmitate oxidation to tricarboxylic acid (TCA) cycle activity in the various perfusion groups is shown in Fig. 2. The contribution of glucose and palmitate to TCA cycle activity was calculated based on the contribution of two acetyl-CoA molecules to the TCA cycle for every molecule of glucose oxidized and eight acetyl-CoA molecules for every molecule of palmitate oxidized. In the control hearts perfused with low fat/low glucose, glucose oxidation provided 29% of the acetyl-CoA for TCA cycle activity. Insulin increased the contribution of glucose oxidation to TCA cycle activity to 53%. In diabetic rat hearts perfused with low fat/low glucose, the contribution of glucose oxidation to TCA cycle activity was <3%, regardless of whether insulin was present or absent from the perfusate.

View larger version (23K): [in this window] [in a new window]   Fig. 2.   Relative contribution of glucose oxidation and palmitate oxidation to tricarboxylic acid (TCA) cycle activity in isolated working control and diabetic rats. ATP production from glucose and palmitate oxidation was calculated from the steady-state rates shown in Fig. 1 using values of 2 moles of acetyl-CoA produced per mole of glucose oxidized (open bars) and 8 moles of acetyl-CoA produced per mole of palmitate oxidized (filled bars).

ACC and AMPK activity. Because fatty acid oxidation rates were the predominant source of energy in the diabetic rat hearts, regardless of the perfusion conditions used, we investigated what effects diabetes had on the subcellular control mechanisms of fatty acid oxidation. ACC activity in control and diabetic hearts frozen at the end of perfusion is shown in Table 3. Basal ACC activity (0 mM citrate) was similar in control and diabetic rat hearts and was not affected by altering the substrate conditions. Insulin stimulated ACC activity only when control hearts were perfused with diabetic buffer (high fat/high glucose). Because citrate has been shown to increase the activity of the phosphorylated and inhibited form of ACC, we also measured ACC activity in the presence of 10 mM citrate. Citrate stimulated ACC activity by 2.5- to 4-fold in each group, but no significant difference was observed between control and diabetic rats in any of the perfusion groups.

MCD activity and expression. Because MCD is also an important regulator of fatty acid oxidation, we measured MCD activity in the control and diabetic rat hearts. As shown in Fig. 3, MCD activity was significantly increased in diabetic rat hearts compared with control hearts in all perfusion groups. MCD activity was not affected in either control or diabetic hearts by altering glucose and fatty acid concentration. Insulin did not change MCD activities in control hearts and diabetic hearts perfused with low-fat/low-glucose buffer but stimulated MCD activity in diabetic hearts perfused with high-fat/high-glucose buffer.

View larger version (22K): [in this window] [in a new window]   Fig. 3.   Malonyl-CoA decarboxylase (MCD) activity in control and diabetic rat hearts. Values are means ± SE; n = 6-10 hearts in each group. Hearts were perfused with Krebs-Henseleit buffer containing either 0.4 mM palmitate and 5 mM glucose (low-fat/low-glucose buffer) or 1.2 mM palmitate and 20 mM glucose (high-fat/high-glucose buffer) in the presence or absence of 100 µU/ml insulin. *Significantly different from control hearts perfused with low-fat/low-glucose buffer in the absence of insulin.

View larger version (19K): [in this window] [in a new window]   Fig. 4.   Levels of MCD protein in control and diabetic rat hearts. Hearts from control and diabetic rats were subjected to SDS-PAGE and visualized by immunoblot analysis using an anti-MCD antibody, as described in MATERIALS AND METHODS. A: representative samples from control and diabetic rats. B: averaged relative content of MCD in 5 control and 5 diabetic rats. Values represent means ± SE. *Significantly different from control (P < 0.05).

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

In uncontrolled diabetes, fatty acid oxidation rates are elevated and account for almost all of the ATP production in the heart (12, 29). This was confirmed in this study, where we demonstrated that over 95% of the acetyl-CoA for the TCA cycle activity in the diabetic rat hearts originated from the oxidation of fatty acids, regardless of the perfusion conditions used. We also found that perfusion of diabetic rat hearts under normal conditions (low concentrations of fatty acid and glucose and in the presence of insulin) did not result in any major change in substrate preference compared with hearts perfused under diabetic conditions (high concentrations of fatty acids and glucose and in the absence of insulin). These results suggest that subcellular changes rather than alterations in energy substrate and insulin levels play an important role in high rates of fatty acid oxidation seen in diabetic hearts. Two important additional conclusions can also be made from our experimental data. First, despite their important role in regulating fatty acid oxidation, neither changes in the activity of ACC nor AMPK can explain the high fatty acid oxidation rates in the diabetic rat heart. Second, MCD activity and protein expression are increased in the diabetic rat heart. Whether this observed increase in MCD activity and expression is due directly to diabetes or to a chronic compensation to increased extracellular substrate secondary to diabetes is not known. However, our data suggest that it is an important contributor to the high fatty acid oxidation rates observed in these hearts.

The effects of diabetes and insulin on control of glucose uptake and metabolism have been investigated extensively. Diabetes results in a decrease in glucose uptake through GLUT-4 transporters, a decrease in glycolysis, and a decrease in glucose oxidation (see Ref. 29 for review). Because diabetes dramatically decreases glucose metabolism in the heart, it has to be considered that the high fatty acid oxidation rates in the diabetic heart were occurring simply due to an inability of the heart to metabolize glucose, as opposed to direct diabetes-induced changes in fatty acid oxidation. This possibility is unlikely for a number of reasons. The first reason is that addition of insulin to control hearts (which presumably have normal insulin-dependent GLUT-4 transport) did not have any significant effect on fatty acid oxidation rates regardless of whether hearts were perfused with high or low concentrations of fatty acids (Fig. 1B). The second reason is that glucose oxidation rates in the heart are much more sensitive to changes in fatty acid oxidation rates than vice versa. For instance, although insulin is capable of stimulating glucose oxidation in control hearts perfused in the presence of low fatty acid concentrations, the ability of insulin to stimulate glucose oxidation was lost at high levels of fatty acids (Fig. 1A). Finally, previous studies have shown that glucose oxidation rates (which provide the majority of ATP derived from glucose metabolism) are inhibited to a much greater extent than glycolysis in the diabetic rat heart (7, 15, 26). This is due to the fact that the rate of glucose oxidation in the hearts appears to be primarily dependent on the rates of fatty acid oxidation (22, 28). This contrasts glycolysis, in which insulin-stimulated glucose uptake is a key determinant of flux through this pathway. Combined, these data suggest that the overreliance of the diabetic rat hearts on fatty acid oxidation is not occurring secondary to a decrease in glucose metabolism but rather due to direct alterations in the control of fatty acid oxidation.

A key enzyme involved in myocardial fatty acid oxidation is CPT 1, which is the rate-limiting enzyme for mitochondrial fatty acid uptake. Heart mitochondria contain an isoform of CPT 1 that is slightly smaller than and immunologically distinct from liver CPT 1 (9, 33), and the enzyme in the heart mitochondria is 50-100 times more sensitive to malonyl-CoA than CPT 1 in the liver of the rat (27). Although the sensitivity of hepatic CPT 1 to malonyl-CoA inhibition decreases in diabetic animals, cardiac CPT 1 sensitivity to malonyl-CoA inhibition is not changed in diabetic animals (4). Recent evidence suggests that in the heart it is the actual control of malonyl-CoA levels that may be a key factor in regulating fatty acid oxidation. However, one problem with addressing this issue is that measurement of absolute levels of malonyl-CoA is not an accurate reflection of the malonyl-CoA accessible to CPT 1. For instance, although malonyl-CoA inhibits cardiac CPT 1 in the 30-60 nM range, if all malonyl-CoA in the heart were cytoplasmic the levels of malonyl-CoA would be in the low micromolar range. This suggests that all myocardial malonyl-CoA is not accessible to CPT 1. Unfortunately, it is technically very difficult to accurately measure cellular distribution of malonyl-CoA in our isolated hearts. As shown in Table 4, absolute tissue levels of malonyl-CoA were not different between control and diabetic rat hearts. However, this does not necessarily mean that malonyl-CoA levels accessible to CPT 1 were not decreased. Because this problem could not be readily resolved, we measured the activity of key enzymes known to regulate malonyl-CoA synthesis (AMPK, ACC, and MCD), since our previous studies have suggested that changes in the activity of these enzymes are paralleled by changes in myocardial fatty acid oxidation rates (5, 10, 11).

In the heart, malonyl-CoA is synthesized by cytoplasmic ACC (13, 14). Two isoforms of ACC are expressed in the heart, a 280-kDa isoform (which predominates) and a 265-kDa isoform (30). Previous studies have shown that decreases in heart ACC activity are associated with increases in fatty acid oxidation rates (5, 10, 20, 21, 24). An important kinase that phosphorylates and inhibits ACC is AMPK (10, 11, 21). In this study, myocardial ACC and AMPK activities in diabetic animals were comparable with those seen in the control hearts, indicating that neither changes in ACC activity nor AMPK activity were responsible for the subcellular alterations that led diabetic hearts to be almost entirely dependent on fatty acid oxidation as an energy source. The results of the present study also showed that insulin stimulated ACC activity only when control hearts were perfused with high-fat/high-glucose buffer and did not alter ACC activity in other conditions. AMPK activity was not changed in either the control or diabetic rat hearts. In the newborn heart, we showed that ACC activity and malonyl-CoA levels in 1-day-old rabbit hearts perfused with insulin were significantly elevated compared with hearts perfused in the absence of insulin (20). Earlier studies in hepatoma cells have also shown that insulin activation of ACC is accompanied by AMPK inhibition (32) and that insulin increases malonyl-CoA production in Langendorff perfused rat hearts (1). The precise reasons for these different results are not clear but warrant further investigation.

Although many investigations have focused on the mechanisms that regulate malonyl-CoA synthesis, few studies have addressed how malonyl-CoA is degraded in the heart. We recently demonstrated that the heart contains an active MCD, which can decarboxylate malonyl-CoA to acetyl-CoA (5). An increase in MCD activity also results in an acceleration of fatty acid oxidation in the newborn heart, whereas high MCD activity contributes to the high fatty acid oxidation rates that occur during reperfusion of ischemic hearts (5). In the present study, a significant increase in MCD activity was observed in the diabetic rat hearts, regardless of the perfusion conditions used. This was accompanied by an increase in the expression of the MCD protein. As a result, we speculate that this increase in MCD activity contributes to the high rates of fatty acid oxidation observed in the diabetic rat hearts. The actual contribution of high MCD activity to the high fatty acid oxidation rates cannot be determined with certainty, due to the lack of available inhibitors of MCD.

In addition to exogenous fatty acid supply, endogenous triacylglycerol stores are also a key source of fatty acids for mitochondrial fatty acid -oxidation (25). Although we have previously shown that myocardial triacylglycerol turnover in streptozotocin diabetic rats is not accelerated compared with control rats perfused with high levels of fatty acids, levels of triacylglycerol in these diabetic rats are elevated (26). As a result, it is possible that the high MCD activity/expression observed in this study also increases endogenous triacylglycerol degradation. In contrast, lower MCD rates and fatty acid oxidation rates observed in control rats may redirect fatty acids away from -oxidation and toward triacylglycerol synthesis. However, the role of MCD in myocardial triacylglycerol turnover has yet to be determined.

In summary, our data demonstrate that an increased MCD activity and expression occur in diabetic rat hearts, which may contribute to the high fatty acid oxidation rates seen in these hearts. ACC and AMPK activity of diabetic hearts was not different from that of control, and altered circulating levels of substrates and insulin in diabetics are not responsible for accelerated fatty acid utilization. These results suggest that elevated MCD activity/expression may have an important contribution to the high fatty acid oxidation rates in the streptozotocin-induced diabetic hearts.