The Antianginal Drug Trimetazidine Shifts Cardiac Energy Metabolism From Fatty Acid Oxidation to Glucose Oxidation by Inhibiting Mitochondrial Long-Chain 3-Ketoacyl Coenzyme A Thiolase
Abstract—Trimetazidine is a clinically effective antianginal agent that has no negative inotropic or vasodilator properties. Although it is thought to have direct cytoprotective actions on the myocardium, the mechanism(s) by which this occurs is as yet undefined. In this study, we determined what effects trimetazidine has on both fatty acid and glucose metabolism in isolated working rat hearts and on the activities of various enzymes involved in fatty acid oxidation. Hearts were perfused with Krebs-Henseleit solution containing 100 μU/mL insulin, 3% albumin, 5 mmol/L glucose, and fatty acids of different chain lengths. Both glucose and fatty acids were appropriately radiolabeled with either 3H or 14C for measurement of glycolysis, glucose oxidation, and fatty acid oxidation. Trimetazidine had no effect on myocardial oxygen consumption or cardiac work under any aerobic perfusion condition used. In hearts perfused with 5 mmol/L glucose and 0.4 mmol/L palmitate, trimetazidine decreased the rate of palmitate oxidation from 488±24 to 408±15 nmol · g dry weight−1 · minute−1 (P<0.05), whereas it increased rates of glucose oxidation from 1889±119 to 2378±166 nmol · g dry weight−1 · minute−1 (P<0.05). In hearts subjected to low-flow ischemia, trimetazidine resulted in a 210% increase in glucose oxidation rates. In both aerobic and ischemic hearts, glycolytic rates were unaltered by trimetazidine. The effects of trimetazidine on glucose oxidation were accompanied by a 37% increase in the active form of pyruvate dehydrogenase, the rate-limiting enzyme for glucose oxidation. No effect of trimetazidine was observed on glycolysis, glucose oxidation, fatty acid oxidation, or active pyruvate dehydrogenase when palmitate was substituted with 0.8 mmol/L octanoate or 1.6 mmol/L butyrate, suggesting that trimetazidine directly inhibits long-chain fatty acid oxidation. This reduction in fatty acid oxidation was accompanied by a significant decrease in the activity of the long-chain isoform of the last enzyme involved in fatty acid β-oxidation, 3-ketoacyl coenzyme A (CoA) thiolase activity (IC50 of 75 nmol/L). In contrast, concentrations of trimetazidine in excess of 10 and 100 μmol/L were needed to inhibit the medium- and short-chain forms of 3-ketoacyl CoA thiolase, respectively. Previous studies have shown that inhibition of fatty acid oxidation and stimulation of glucose oxidation can protect the ischemic heart. Therefore, our data suggest that the antianginal effects of trimetazidine may occur because of an inhibition of long-chain 3-ketoacyl CoA thiolase activity, which results in a reduction in fatty acid oxidation and a stimulation of glucose oxidation.
Trimetazidine (1-[2,3,4-trimethoxybenzyl] piperazine dihydrochloride) is a clinically effective antianginal agent. In several double-blind trials, trimetazidine was shown to improve the ergometric exercise capacity and total work output of patients with effort angina,1 2 to reduce attack frequency and nitroglycerin requirement in patients with chronic stable angina,3 and to increase effort tolerance in angina patients.4 Multicenter trials of trimetazidine by a European collaborative working group have demonstrated that the antianginal efficacy of trimetazidine is equivalent to that of propranolol but does not reduce cardiac rate-pressure product or coronary blood flow.5 A similar lack of alteration of hemodynamic parameters with comparable antianginal effect was noted in comparison with nifedipine.6 In combination with diltiazem, additive antianginal effects were also observed.7 8 Trimetazidine has thus been considered unique among antianginal agents for its lack of vasodilator activity and has been termed a “cellular anti-ischemic agent (p 292).”9 It is presently clinically used throughout Europe and in >80 countries worldwide.
Despite these clinical successes, the understanding of trimetazidine’s mechanism of action remains incomplete. It is proposed to act by directly improving myocardial energy metabolism, resulting in cytoprotective effects in several models of myocardial ischemia.9 10 11 How could trimetazidine improve myocardial energy balance? Switching energy substrate preference from fatty acid oxidation to glucose oxidation is one possible means of improving cardiac function during ischemia and reperfusion, as well as improving both cardiac energetics and cardiac efficiency (see References 12 through 15 for reviews). A study by Fantini et al16 has shown in isolated rat heart mitochondria that trimetazidine has a potent inhibitory effect on palmitoylcarnitine oxidation, with no significant effect on pyruvate oxidation. This indirect evidence suggests that trimetazidine may act to inhibit fatty acid oxidation in the heart. Consistent with this, we have recently shown that another structurally similar piperazine derivative, ranolazine, also increases glucose oxidation in isolated working rat hearts, secondary to an inhibition of fatty acid oxidation.17
This indirect evidence suggests that trimetazidine may exert its antianginal effects by inhibiting fatty acid oxidation. High fatty acid oxidation rates are detrimental in the setting of ischemia and reperfusion because of an inhibition of glucose oxidation.18 19 This leads to an increase in proton production due to an uncoupling of glycolysis from glucose oxidation.19 20 This increase in proton production has the potential to accelerate sodium and calcium overload in the heart,21 resulting in an exacerbation of ischemic injury and a decrease in cardiac efficiency during reperfusion.14 15 19 20 Despite the observation that trimetazidine both improves cardiac energetics10 11 16 and attenuates proton and sodium accumulation during and after ischemia,22 it has not been directly determined what effect trimetazidine has on myocardial energy metabolism.
The purpose of this study was to determine whether trimetazidine exerts a direct effect on glycolysis, glucose oxidation, or fatty acid oxidation in the heart and to determine at what site trimetazidine alters energy metabolism in the heart. We demonstrate that trimetazidine potently inhibits long-chain 3-ketoacyl coenzyme A (CoA) thiolase in the heart, resulting in a reduction in fatty acid oxidation and an increase in glucose oxidation. This stimulation of glucose oxidation may explain the cardioprotective effect of trimetazidine seen in fatty acid–perfused ischemic rat hearts.
Materials and Methods
Isolated working hearts from male Sprague-Dawley rats were subjected to a 60-minute aerobic perfusion period with a modified Krebs-Henseleit solution containing 5 mmol/L glucose; 100 μU/mL insulin; 3% BSA; and (in mmol/L) palmitate (16-carbon) 0.4, octanoate (8-carbon) 0.8, or butyrate (4-carbon) 1.6. Working hearts were used in these studies to approximate the metabolic demand of the heart seen in vivo.
An additional series of hearts perfused with 5 mmol/L glucose, 100 μU/mL insulin and 0.4 mmol/L palmitate were subjected to a 30-minute perfusion, after which they were switched to the Langendorff mode and subjected to a 60-minute period of low-flow ischemia (coronary flow=0.5 mL/min). Hearts were switched to the Langendorff mode during ischemia to obtain a constant coronary flow rate during ischemia.
Another series of hearts were perfused with Krebs-Henseleit perfusate containing 11 mmol/L glucose, 100 μU/mL insulin, and 1.2 mmol/L palmitate.
Spontaneously beating hearts were perfused at an 11.5–mm Hg left atrial preload and an 80–mm Hg aortic afterload. Heart rate, peak systolic pressure, developed pressure, cardiac output, aortic flow, coronary flow, cardiac work, and O2 consumption were measured as described previously.18 19 23
Trimetazidine, when present, was added 5 minutes into the perfusion period at the concentrations indicated in the tables and figures.
Glycolysis and glucose oxidation rates were determined simultaneously by the quantitative collection of 3H2O and 14CO2 produced by hearts perfused with buffer containing [5-3H/U-14]glucose.19 Rates of fatty acid oxidation were assessed during identical parallel heart perfusions using [1-14C]palmitate, [1-14C]octanoate, or [1-14C]butyrate. Palmitate, octanoate, or butyrate oxidation was determined by quantitatively collecting 14CO2 production as described above for glucose oxidation.
The rate of acetyl CoA production calculated from tricarboxylic acid (TCA)-cycle activity was as described previously.19 TCA-cycle efficiency was defined as the ratio of cardiac work to TCA-cycle acetyl CoA produced.
Analysis of active and total pyruvate dehydrogenase (PDHa and PDHt, respectively) was as described previously.23 24 Activities of long-chain, medium-chain, and short-chain acyl CoA dehydrogenase in isolated rat heart mitochondria were measured as described by Grimbert et al.25 The long-, medium- and short-chain enoyl CoA hydratase activities were assayed as described by Wanders et al.26 Long-, medium, and short-chain l-3-hydroxyacyl CoA dehydrogenase and 3-ketoacyl CoA thiolase were measured in membrane and soluble mitochondrial fractions as described by Venizelos et al.27
CoA esters were extracted from the frozen powdered heart tissue and quantified by high-performance liquid chromatography (HPLC).28 29 Myocardial AMP-activated protein kinase (AMPK), acetyl CoA carboxylase (ACC), malonyl CoA carboxylase, and triacylglycerols were measured as described previously.29 30 31
All data are represented as the group mean±SEM. Data were analyzed using the statistical program Instat. One-way ANOVA was used to compare values among groups. A Tukey-Grammar post hoc test was used to confirm intergroup differences. P<0.05 was considered significant.
An expanded Materials and Methods section is available online at http://www.circresaha.org.
Effect of Trimetazidine on Cardiac Function
Table 1⇓ online (see http://www.circresaha.org) shows the effect of trimetazidine (1 μmol/L) on cardiac function in isolated working rat hearts perfused in the presence of glucose as well as palmitate, octanoate, or butyrate. Regardless of the type of fatty acid substrate used, the presence of 1 μmol/L trimetazidine did not cause any significant change in heart rate, peak systolic pressure, developed pressure, cardiac output, or cardiac work. This confirms earlier studies demonstrating that trimetazidine is not an inotropic agent and demonstrates that any changes in energy substrate metabolism in the heart are not occurring secondary to effects of trimetazidine on cardiac function. Trimetazidine also did not significantly alter O2 consumption in any of the perfusion groups. Trimetazidine also did not have any effect on coronary artery flow either, which argues against a primary vasodilatory effect of this drug in the normoxic setting.
Effects of Trimetazidine on Energy Metabolism
The effect of trimetazidine on glucose and fatty acid utilization rates in the presence of a physiological level of long-chain fatty acid substrate (palmitate 0.4 mmol/L) is demonstrated in Figure 1A⇓ through 1C. Rates of glycolysis were unaltered by trimetazidine, although there was a significant increase in glucose oxidation rates. In these hearts, glycolytic rates were significantly greater than glucose oxidation rates (Figure 1A⇓ and 1B⇓). As a result, the selective increase in glucose oxidation rates resulted in an improved coupling between glycolysis and glucose oxidation. The increase in glucose oxidation rates in the trimetazidine hearts was accompanied by a significant decrease in palmitate oxidation rates (Figure 1C⇓).
The effects of trimetazidine on glycolysis and glucose oxidation in hearts subjected to low-flow ischemia (coronary flow=0.5 mL/min) are shown in Figure 2⇓. During ischemia, glycolytic rates in control hearts were only slightly lower than rates observed under aerobic conditions (Figure 1A⇑). However, as expected, a substantial decrease in glucose oxidation rates was observed during ischemia (Figures 2⇓ and 1B⇑). Addition of trimetazidine had no effect on glycolytic rates but did result in a significant increase in glucose oxidation rates.
To determine whether the inhibitory effects of trimetazidine on fatty acid oxidation were specific for long-chain fatty acids, energy metabolism was also measured in hearts perfused with octanoate (an 8-carbon fatty acid) (Figure 3⇓) and butyrate (a 4-carbon fatty acid) (Figure 4⇓). In octanoate-perfused hearts, trimetazidine did not have any significant effect on glycolysis, glucose oxidation, or octanoate oxidation. Similarly, in the presence of butyrate, trimetazidine was without effect on glycolysis, glucose oxidation, or butyrate oxidation. This suggests that trimetazidine specifically inhibits long-chain fatty acid oxidation and that the stimulation of glucose oxidation does not occur in the presence of shorter-chain fatty acids.
The effects of trimetazidine on cardiac work/TCA-cycle activity were also calculated to determine whether trimetazidine alters TCA-cycle efficiency in aerobically perfused hearts. As shown in Figure 1⇑ online (see http://www.circresaha.org), no effect of trimetazidine on cardiac work/TCA-cycle activity was observed regardless of whether hearts were perfused with palmitate, octanoate, or butyrate. This suggests that the decrease in palmitate oxidation by trimetazidine is accompanied by a parallel increase in glucose oxidation.
We also determined whether trimetazidine was able to stimulate glucose oxidation in hearts perfused with high levels of fatty acids (Table 1⇑). As expected, rates of glucose oxidation in untreated hearts were much lower in hearts perfused with 1.2 mmol/L palmitate (347±58 nmol · g dry weight−1 · minute−1) compared with hearts perfused with 0.4 mmol/L palmitate (1889±119 nmol · g dry weight−1 · minute−1) (Figure 1B⇑). In contrast, rates of glycolysis were not affected by fatty acid concentration, confirming that fatty acids are much more potent inhibitors of glucose oxidation than of glycolysis. As shown in Table 1⇑, even with glucose oxidation rates inhibited by high levels of fatty acids, trimetazidine was able to significantly stimulate glucose oxidation. A significant stimulation of glucose oxidation was observed in the presence of 0.1 μmol/L trimetazidine. A maximal stimulation of glucose oxidation was observed in the presence of 1 μmol/L trimetazidine.
The heart contains a sizeable triacylglycerol pool that can contribute as a source of fatty acids for β-oxidation. We therefore measured triacylglycerol levels in the palmitate-, octanoate-, and butyrate-perfused hearts at the end of the 60-minute perfusion period. Trimetazidine had no effect on triacylglycerol levels in the palmitate-perfused hearts (145.3±15.6 versus 150.6±38.4 μmol fatty acid/g dry weight, in control and trimetazidine-treated hearts, respectively; P=NS). Under control conditions, triacylglycerol levels in octanoate- and butyrate-perfused hearts (88.1±10.2 and 99.2±6.3 μmol fatty acid/g dry weight, respectively) were lower than levels seen in palmitate-perfused hearts (P<0.05). The oxidation of unlabeled fatty acids released from triacylglycerol hydrolyzed from these groups would explain the slightly higher cardiac work/TCA-cycle activity observed in the octanoate- and butyrate-perfused hearts compared with the palmitate-perfused hearts (Figure 1⇑ online, see http://www.circresaha.org). However, similar to the palmitate-perfused hearts, trimetazidine had no effect on triacylglycerol levels in the octanoate- and butyrate-perfused hearts (83.4±6.6 and 104.9±17.2 μmol fatty acid/g dry weight, respectively; P=NS compared with the appropriate control perfusion group).
Effects of Trimetazidine on Pyruvate Dehydrogenase (PDH) Activity
PDH is the rate-limiting step for glucose oxidation and is inhibited when fatty acid oxidation rates are high (see Reference 14 for review). We therefore measured PDH activity in hearts perfused with 5 mmol/L glucose and 0.4 mmol/L palmitate (Table 2⇓). In control hearts, the active unphosphorylated form of the enzyme (PDHa) was 18% as active as PDHt. These values are expected in hearts perfused with normal levels of fatty acids. In hearts perfused with 1 μmol/L trimetazidine, a significant increase in PDHa was observed, with no difference in PDHt. As a result, a significant increase in the active-to-total ratio was observed. In hearts perfused with octanoate, in which no increase in glucose oxidation was observed, no significant difference in the PDHa (4.12±0.48 and 5.16±0.56 μmol acetyl CoA · g dry weight−1 · minute−1, in control and trimetazidine-treated hearts, respectively; n=10 in each group) or PDHt was observed (23.9±2.1 and 22.4±2.6 μmol acetyl CoA · g dry weight−1 · minute−1 in control and trimetazidine-treated hearts, respectively; n=10 in each group).
To determine whether the effects of trimetazidine were due to direct actions on PDH itself, a large concentration of trimetazidine (100 μmol/L) was added directly to the incubation medium used to measure PDHa and PDHt. As shown in Table 2⇑ online (see http://www.circresaha.org), trimetazidine had no direct effect on either PDHa or PDHt, suggesting that the stimulation of PDHa seen in Table 2⇑ online was an indirect effect on trimetazidine and not due to a direct stimulation of PDHa activity.
Effect of Trimetazidine on Levels of Malonyl CoA and the Enzymes That Control Malonyl CoA Levels
Malonyl CoA is a potent inhibitor of fatty acid oxidation due to inhibition of carnitine palmitoyltransferase (CPT) 1, the rate-limiting step in fatty acid oxidation. We therefore determined what effect trimetazidine had on malonyl CoA levels, because measurements of energy metabolism suggest that trimetazidine may directly inhibit fatty acid oxidation. The effect of trimetazidine on malonyl CoA levels was measured in hearts perfused with palmitate, octanoate, or butyrate. As shown in Table 3⇓ online (see http://www.circresaha.org), trimetazidine had no effect on malonyl CoA levels in any of these 3 perfusion groups.
Malonyl CoA in the heart is synthesized by ACC and degraded by malonyl CoA decarboxylase (MCD).31 We therefore determined what effect trimetazidine had on enzyme activity in the hearts perfused with palmitate, as well as what effect trimetazidine had directly on enzyme activity. Addition of trimetazidine to palmitate-perfused hearts did not result in any change in either ACC activity (8.23±0.4 and 6.43±0.35 nmol · mg protein−1 · minute−1 [P=NS] in control and trimetazidine-treated hearts, respectively; n=9 in each group). Trimetazidine also had no effect on citrate-stimulatable ACC activity. It also had no effect on MCD activity (8.01±0.46 and 6.87±0.29 μmol · mg protein−1 · minute−l [P=NS], in control and trimetazidine-treated hearts, respectively; n=9 in each group). Although the cellular location of MCD has yet to be established (ie, some is probably intramitochondrial), our data suggest that trimetazidine does not alter fatty acid oxidation by altering MCD activity. Because ACC is also inhibited by AMPK, we also determined what effect trimetazidine had on AMPK activity in the palmitate-perfused hearts. Similar to ACC and MCD, no significant effect of trimetazidine on AMPK activity was observed (852±111 and 1022±191 pmol · mg protein−1 · minute−l [P=NS] in control and trimetazidine-treated hearts, respectively, n=9 in each group).
Direct addition of 10 μmol/L trimetazidine to the incubation medium was also without effect on the activity of ACC, MCD, or AMPK (data not shown).
Effect of Trimetazidine on Myocardial CoA Ester Levels
The effects of trimetazidine on other measured short-chain CoA esters in the palmitate-, octanoate-, and butyrate-perfused hearts are also shown in Table 3⇑ online (see http://www.circresaha.org). No effect of trimetazidine was observed on any of the measured CoA esters. The lack of effect on succinyl CoA levels suggests that, despite the inhibition of fatty acid oxidation in palmitate-perfused hearts, the supply of carbon for the TCA cycle was not compromised.
The effects of trimetazidine on levels of long-chain CoA, acetyl CoA, and free CoA are shown in Table 4 online (see http://www.circresaha.org). As expected, levels of long-chain acyl CoA were highest in hearts perfused with palmitate, compared with either octanoate- or butyrate-perfused hearts. However, trimetazidine did not affect long-chain acyl CoA levels in any perfusion group. Similarly, no effect of trimetazidine on acetyl CoA levels was observed in any of the perfusion groups. Trimetazidine did result in a significant increase in free CoA in hearts perfused with palmitate, which would be expected if trimetazidine directly inhibited long-chain fatty acid β-oxidation. However, the ratio of acetyl CoA/CoA was not significantly affected by trimetazidine in any of the perfusion groups.
Effect of Trimetazidine on the Enzymes of Fatty Acid β-Oxidation
Because reduction of fatty acid oxidation by trimetazidine could not be explained by alterations in malonyl CoA control of mitochondrial fatty acid uptake, we directly measured the effects of trimetazidine on the enzymes of mitochondrial fatty acid β-oxidation. Table 3⇑ summarizes the activities of long-, medium- and short-chain isoforms of acyl CoA dehydrogenase, enoyl CoA dehydrogenase, and l-3-hydroxyacyl CoA dehydrogenase activity. The presence of 10 or 75 μmol/L trimetazidine did not have any significant effect on any of the isoforms for the first 3 enzymes involved in mitochondrial fatty acid β-oxidation. Concentration effects of trimetazidine on enzyme activities were also measured, which confirmed that trimetazidine did not alter the activity of any of these enzymes (data not shown).
The effect of trimetazidine on the long-, medium, and short-chain isoform of 3-ketoacyl CoA thiolase is shown in Figure 5⇓. Trimetazidine caused a potent inhibition of long-chain 3-ketoacyl CoA thiolase, with an IC50 of ≈75 nmol/L. In contrast, a concentration of 10 μmol/L trimetazidine was necessary to significantly inhibit the medium-chain 3-ketoacyl CoA thiolase, and concentrations in excess of 100 μmol/L were necessary to inhibit the short-chain 3-ketoacyl CoA thiolase. In these experiments, 4-pentenoic acid, a nonselective inhibitor of 3-ketoacyl CoA thiolase, was used as a positive inhibitor control. As expected, 4-pentenoic acid inhibited the enzyme activities of all 3 isoforms of 3-ketoacyl CoA thiolase.
Although the clinical efficacy of trimetazidine as an anti-ischemic agent is undisputed, its cellular mechanism of action has not been previously delineated. In this study, we demonstrate that trimetazidine directly inhibits fatty acid oxidation in the heart, secondary to an inhibition of mitochondrial long-chain 3-ketoacyl CoA thiolase. The selective inhibition of the long-chain 3-ketoacyl CoA thiolase is consistent with a specific inhibition by trimetazidine of long-chain fatty acid oxidation and not medium- or short-chain fatty acids. This inhibition of long-chain fatty acid oxidation is accompanied by an increase in PDH activity, resulting in a significant increase in glucose oxidation rates. The increase in glucose oxidation is seen regardless of whether hearts are perfused with low or high concentrations of fatty acids and is also observed in hearts subjected to low-flow ischemia. Furthermore, trimetazidine is without effect on glycolytic rates, which will improve the coupling between glycolysis and glucose oxidation and a decrease in proton production from glucose metabolism. Because increasing glucose oxidation and decreasing proton production from glucose metabolism is cardioprotective (see References 14 and 15 for reviews), we hypothesize that reduction in fatty acid oxidation is the mechanism by which trimetazidine exerts its antianginal effects. Our data also implicate inhibition of long-chain 3-ketoacyl CoA thiolase as the site of action for the beneficial anti-ischemic effects of trimetazidine. It also confirms that reduction in fatty acid oxidation is a novel approach to treating ischemic heart disease.
Trimetazidine and Myocardial Fatty Acid Oxidation
The demonstration that trimetazidine can inhibit fatty acid oxidation in palmitate-perfused hearts is consistent with the previous studies of Fantini et al16 using isolated mitochondrial preparations. These authors demonstrated that trimetazidine could markedly reduce in vitro mitochondrial respiratory activity but that this effect is dependent on the substrate offered to the mitochondria; ie, palmitoylcarnitine oxidation is markedly reduced, but not that of pyruvate, glutamate, or citrate.16 This suggests that trimetazidine is not inhibiting either TCA-cycle activity or mitochondrial oxidative phosphorylation, but rather is directly inhibiting fatty acid β-oxidation. This is supported by our data, in which we demonstrate that trimetazidine does not alter overall oxygen consumption in aerobically perfused hearts (Table 1⇑ online, see http://www.circresaha.org), nor does it alter TCA-cycle activity or cardiac work/TCA-cycle activity (Figure 1⇑ online, see http://www.circresaha.org). Rather, the effects of trimetazidine are associated with a switch in the source of acetyl CoA for the TCA cycle from fatty acid β-oxidation to glucose oxidation. This is not due to a direct stimulation of PDH, but rather an indirect stimulation of PDH secondary to a reduction in fatty acid oxidation. Because fatty acid β-oxidation rates are a key determinant of PDH activity,32 trimetazidine reduction in fatty acid oxidation was accompanied by an increase in glucose oxidation, with no overall decrease in the levels of TCA-cycle intermediates (Tables 3 and 4 online, see http://www.circresaha.org).
Mitochondrial uptake of fatty acids is another important site at which myocardial fatty acid oxidation can be inhibited. CPT 1 is an important enzyme in this process and is potently inhibited by malonyl CoA. Because long-chain fatty acid uptake by the mitochondria, and not medium- or short-chain fatty acid uptake, is dependent on CPT 1, it is possible that the effects of trimetazidine on fatty acid oxidation may be occurring secondary to an inhibition of CPT 1 activity. However, we believe this is unlikely for a number of reasons, as follows. (1) Previous studies have shown that trimetazidine does not directly inhibit CPT 1 activity.33 (2) Inhibition of mitochondrial respiration of palmitoylcarnitine would not occur if trimetazidine were acting at the level of CPT 1.16 (3) CPT 1 inhibitors result in a decrease in long-chain acyl CoA levels, presumably because of a decrease in mitochondrial long-chain acyl CoA levels.34 As shown in our studies, trimetazidine did not lower long-chain acyl CoA levels (Table 3⇑ online [see http://www.circresaha.org]), which is consistent with a site of inhibition lower in the fatty acid β-oxidation pathway. (4) Trimetazidine did not alter the levels of malonyl CoA under any perfusion condition, nor did it alter the activity of any of the enzymes involved in the regulation of malonyl CoA synthesis or degradation (ACC, MCD, or AMPK).
Combined, these data suggest that trimetazidine inhibits fatty acid oxidation directly in the β-oxidative pathway.
The β-oxidation of fatty acids utilizes 4 enzymes, including acyl CoA dehydrogenase, enoyl CoA hydratase, l-3-hydroxyacyl CoA dehydrogenase, and 3-ketoacyl CoA thiolase. Each of these enzymes exists as a number of different isoforms, with each isoform having a different specificity for fatty acyl-chain length (see Reference 35 for review). This allows complete oxidation of fatty acids as successive cycles shorten the fatty acyl esters. As shown in Table 3⇑, trimetazidine did not inhibit any isoform of the first 3 β-oxidative enzymes, but was a potent inhibitor of long-chain 3-ketoacyl CoA thiolase (Figure 5⇑), the enzyme catalyzing the last step in β-oxidation. The concentration of trimetazidine necessary to inhibit long-chain 3-ketoacyl CoA thiolase is consistent with the concentration of trimetazidine necessary to stimulate glucose oxidation in hearts perfused with high levels of fatty acids (Table 1⇑) and is compatible with the clinically relevant plasma concentrations seen in patients taking trimetazidine for the treatment of angina.36 Although trimetazidine also inhibited the medium- and short-chain isoforms of 3-ketoacyl CoA thiolase, the concentrations needed were much higher than clinically relevant concentrations. The observation that trimetazidine was ineffective at inhibiting either octanoate or butyrate oxidation is also consistent with an inhibition of long-chain 3-ketoacyl CoA thiolase, given that both of these substrates enter the β-oxidative pathway at a level beyond long-chain 3-ketoacyl CoA thiolase. The selective increase in free CoA by trimetazidine in palmitate-perfused hearts is also consistent with inhibition of long-chain 3-ketoacyl CoA thiolase.
In octanoate-perfused hearts, trimetazidine did result in a small, nonsignificant decrease in octanoate oxidation, as well as a small, nonsignificant increase in PDHa. A possible explanation for this may be that the long-chain 3-ketoacyl CoA thiolase may be metabolizing some of the octanoate. Although the long-chain 3-ketoacyl CoA thiolase prefers fatty acids of 10 carbons or greater in length, it can use 8-carbon–length fatty acids as a substrate (albeit with a lower affinity). Unfortunately, no study to date has clearly defined what proportion of medium-chain 3-ketoacyl CoAs are metabolized by the long- versus medium-chain 3-ketoacyl CoA thiolase in vivo. Therefore, it is possible that inhibition of the long-chain 3-ketoacyl CoA thiolase by trimetazidine could have some effect on overall octanoate oxidation rates (and therefore a minor effect on PDH activity).
Trimetazidine and Glucose Oxidation
Our data also show that stimulation of PDHa and glucose oxidation by trimetazidine does not occur by a direct effect on the PDH complex. This is because trimetazidine had no direct effect on PDHa, and in octanoate-perfused hearts trimetazidine had no effect on PDH activity. The PDH complex that activates and then translocates activated pyruvate into the mitochondrial matrix in the form of acetyl CoA is the rate-limiting step for glucose oxidation by the TCA cycle. PDH activity itself is subject to various control mechanisms, including PDH kinase–mediated phosphorylation, which inactivates the enzyme, and PDH phosphatase–mediated dephosphorylation, which restores the active state.32 Flux of fatty acids through β-oxidation is closely linked to PDHa, because both reduced equivalents and accumulation of acetyl CoA from β-oxidation activate the kinase and decrease PDHa. In contrast, inhibition of fatty acid oxidation has the opposite effect of decreasing kinase activity, increasing PDHa, and stimulating glucose oxidation rates. As a result, trimetazidine reduction in fatty acid oxidation was accompanied by a significant increase in glucose oxidation rates.
Trimetazidine and Cardioprotection During and After Ischemia
Energy metabolism during ischemia and during reperfusion is closely linked to cardiac function, presenting the clinician an arena for possible intervention. Current approaches that are used to manipulate myocardial energy metabolism involve either stimulating glucose metabolism or inhibiting fatty acid metabolism. Numerous experimental studies have demonstrated that stimulation of glucose oxidation both during and after ischemia can benefit the ischemic heart (see References 14 and 15 for reviews). By improving the coupling of glycolysis to glucose oxidation, proton production is decreased, resulting in a decrease in tissue acidosis and an improvement in cardiac efficiency. As a result, selective stimulation of glucose oxidation by trimetazidine may explain the anti-ischemic effects of this agent, possibly by increasing the coupling between glycolysis and glucose oxidation, thereby decreasing proton production.
We propose that trimetazidine acts by reducing fatty acid oxidation, thereby increasing glucose oxidation before and during the ischemic event. This altered metabolic balance will improve the capacity of the ischemic heart to balance the dramatic increase in glycolytic flux with an increase in glucose oxidation. This strategy has been demonstrated to be effective in experimental as well as clinical settings using agents such as dichloroacetate or ranolazine, which stimulate glucose oxidation in the hearts.14 15 17
In summary, we demonstrate that trimetazidine inhibits fatty acid oxidation secondary to an inhibition of long-chain 3-ketoacyl CoA thiolase, resulting in an increase in glucose oxidation. This results in an improved coupling of glycolysis with glucose oxidation, which has previously been shown to protect the ischemic heart. As a result, switching energy substrate preference from fatty acid oxidation to glucose oxidation may explain the antianginal properties of trimetazidine.
This work was funded by a grant from the Institut De Recherches Internationales Servier. P.F.K. is Clinical Fellow of the Heart and Stroke Foundation of Canada and the Alberta Heritage Foundation for Medical Research. G.D.L. is a Heritage Medical Scientist.
- Received December 2, 1999.
- Accepted January 20, 2000.
- © 2000 American Heart Association, Inc.
Sellier P, Audouin P, Payen B, Duong TC, Ourbak P. Ergometric effects of a single administration of trimetazidine. Presse Med. 1986;15:1771–1774.
Passeron J. Effectiveness of trimetazidine in stable effort angina due to chronic coronary insufficiency: a double-blind versus placebo study. Presse Med. 1986;15:1775–1778.
Gallet M. Clinical effectiveness of trimetazidine in stable effort angina: a double blind versus placebo controlled study. Presse Med. 1986;15:1779–1782.
Detry JM, Sellier P, Pennaforte S, Cokkinos D, Dargie H, Mathes P. Trimetazidine: a new concept in the treatment of angina: comparison with propranolol in patients with stable angina. Trimetazidine European Multicenter Study Group. Br J Clin Pharmacol. 1994;3:279–288.
Dalla-Volta S, Maraglino G, Della-Valentina P, Viena P, Desideri A. Comparison of trimetazidine with nifedipene in effort angina: a double-blind, crossover study. Cardiovasc Drugs Ther. 1990;4:853–859.
Manchanda SC, Krishnaswani S. Combination treatment with trimetazidine and diltiazem in stable angina pectoris. Heart. 1997;78:353–357.
Harpey C, Clauser P, Labrid C, Freyria JL, Poirier JP. Trimetazidine, a cellular anti-ischemic agent. Cardiovasc Drug Rev. 1989;6:292–312.
Stanley WC, Lopaschuk GD, Hall JL, McCormack JG. Regulation of myocardial carbohydrate metabolism under normal and ischemic conditions: potential for pharmacological interventions. Cardiovasc Res. 1997;33:243–257.
Fantini E, Demaison L, Sentex E, Grynberg A, Athias P. Some biochemical aspects of the protective effect of Trimetazidine on rat cardiomyocytes during hypoxia and reoxygenation. J Mol Cell Cardiol. 1994;26 949–958.
McCormack JG, Barr RL, Wolff AA, Lopaschuk GD. Ranolazine stimulates glucose oxidation in normoxic, ischemic and reperfused ischemic rat hearts. Circulation. 1996;93:135–142.
Lopaschuk GD, Wambolt RB, Barr RL. An imbalance between glycolysis and glucose oxidation is a possible explanation for the detrimental effects of high levels of fatty acids during aerobic reperfusion of ischemic hearts. J Pharmacol Exp Ther. 1993;264:135–144.
Liu B, el Alaoui-Talibi Z, Clanachan AS, Schulz R, Lopaschuk GD. Uncoupling of contractile function from mitochondrial TCA cycle activity and MVO2 during reperfusion of ischemic hearts. Am J Physiol. 1996;270:H72–H80.
Liu B, Clanachan AS, Schulz R, Lopaschuk GD. Cardiac efficiency is improved after ischemia by altering both the source and fate of protons. Circ Res. 1996;79:940–948.
El Banani H, Bernard M, Baetz D, Cabanes E, Cozzone P, Lucien A, Feuvray D. Changes in nuclear magnetic resonance study of the effects of TMZ on intracellular Na+ and pH during ischemia and reperfusion. J Mol Cell Cardiol.. 1999;31:A53. Abstract.
Collins-Nakai RL, Noseworthy D, Lopaschuk GD. Epinephrine increases ATP production in hearts by preferentially increasing glucose metabolism. Am J Physiol. 1994;267:H1862–H1871.
Grimbert S, Fisch C, Deschamps D, Berson A, Fromenty B, Feldman G, Pessayre D. Effects of female sex hormones on mitochondria: possible role in acute fatty liver of pregnancy. Am J Physiol. 1995;268:G107–G115.
Saddik M, Gamble J, Witters LA, Lopaschuk GD. Acetyl CoA carboxylase regulation of fatty acid oxidation in the heart. J Biol Chem. 1993;268:25836–25845.
Kudo N, Barr A, Barr R, Desai S, Lopaschuk GD. High rates of fatty acid oxidation during reperfusion of ischemic hearts are associated with a decrease in malonyl-CoA levels due to an increase in 5′AMP-activated protein kinase inhibition of acetyl-CoA carboxylase. J Biol Chem. 1995;270:17513–17520.
Hardie DG, Salt IP, Hawley SA, Davies SP. AMP-activated protein kinase: an ultrasensitive system for monitoring cellular energy charge. Biochem J. 1999;338:717–722.
Dyck JRB, Barr AJ, Barr RL, Kolattukudy PE, Lopaschuk GD. Characterization of cardiac malonyl CoA decarboxylase and its putative role in regulating fatty acid oxidation. Am J Physiol. 1998;275:H2122–H2129.
Randle PJ, Hales CN, Garland PB, Newsholme EA. The glucose fatty-acid cycle: its role in insulin sensitivity and the metabolic disturbances of diabetic mellitus. Lancet. 1963;i:785–789.
Lopaschuk GD, Spafford MA, Davies NJ, Wall SR. Glucose and palmitate oxidation in isolated working rat hearts reperfused following a period of transient global ischemia. Circ Res. 1989;66:546–553.