Trimetazidine does not alter metabolic substrate oxidation in cardiac mitochondria of target patient population

Abstract

Background and Purpose

Trimetazidine, known as a metabolic modulator, is an anti‐anginal drug used for treatment of stable coronary artery disease (CAD). It is proposed to act via modulation of cardiac metabolism, shifting the mitochondrial substrate utilization towards carbohydrates, thus increasing the efficiency of ATP production. This mechanism was recently challenged; however, these studies used indirect approaches and animal models, which made their conclusions questionable. The goal of the current study was to assess the effect of trimetazidine on mitochondrial substrate oxidation directly in left ventricular myocardium from CAD patients.

Experimental Approach

Mitochondrial fatty acid (palmitoylcarnitine) and carbohydrate (pyruvate) oxidation were measured in permeabilized left ventricular fibres obtained during coronary artery bypass grafting surgery from CAD patients, which either had trimetazidine included in their therapy (TMZ group) or not (Control).

Key Results

There was no difference between the two groups in the oxidation of either palmitoylcarnitine or pyruvate, and in the ratio of carbohydrate to fatty acid oxidation. Activity and expression of pyruvate dehydrogenase, the key regulator of carbohydrate metabolism, were also not different. Lastly, acute in vitro exposure of myocardial tissue to different concentrations of trimetazidine did not affect myocardial oxidation of fatty acid.

Conclusion and Implications

Using myocardial tissue from CAD patients, we found that trimetazidine (applied chronically in vivo or acutely in vitro) had no effect on cardiac fatty acid and carbohydrate oxidation, suggesting that the clinical effects of trimetazidine are unlikely to be due to its metabolic effects, but rather to an as yet unidentified intracardiac mechanism.

Abbreviations

CABG

coronary artery bypass grafting

CAD

coronary artery disease

CTL

control group of patients

DM

diabetes mellitus

ESC

European Society of Cardiology

ETC

electron transfer chain

FCCP

trifluorocarbonylcyanide phenylhydrazone

LAD

left anterior descending branch

LC 3‐KAT

long‐chain 3‐ketoacyl CoA thiolase

LV

left ventricle

LVEF

left ventricle ejection fraction

PDH

pyruvate dehydrogenase

TMZ

trimetazidine

Tables of Links

TARGETS
Enzymes a	Transporters b
NOS	ATP synthase (F‐type ATPases)

LIGANDS
ADP	Pyruvate
ATP	Ranolazine
Glutamate	Succinate
Malate

These Tables list key protein targets and ligands in this article which are hyperlinked to corresponding entries in http://www.guidetopharmacology.org, the common portal for data from the IUPHAR/BPS Guide to PHARMACOLOGY (Pawson et al., 2014) and are permanently archived in the Concise Guide to PHARMACOLOGY 2015/16 (^a,bAlexander et al., 2015a, 2015b).

Introduction

The modulation of mitochondrial metabolism is a relatively recent paradigm in the treatment of coronary artery disease (CAD), with the ‘metabolic modulator’ trimetazidine (1‐[2,3,4‐trimethoxybenzyl] piperazine dihydrochloride; TMZ) being increasingly used for the treatment of angina symptoms, especially since it was recognized in the European Society of Cardiology (ESC) guidelines for the management of stable CAD (Montalescot et al., 2013). The clinical efficacy of TMZ (Danchin et al., 2011; Fragasso et al., 2013; Vitale et al., 2013) has been attributed to its inhibitory actions on the long‐chain 3‐ketoacyl CoA thiolase (LC 3‐KAT), resulting in inhibition of cardiac β‐oxidation. This, by shifting the metabolism from fatty acid utilization towards oxidation of carbohydrates, would increase the efficiency of ATP production for a given amount of oxygen consumed, thus increasing the energy reserve in a hypoxic myocardium.

This mechanism, which was first described by Kantor et al. (2000) and is currently widely accepted by clinicians, as well as being marketed by the industry, has now been questioned as a result of conclusions from other studies, which did not find a significant inhibitory effect of TMZ on β‐oxidation or found only a mild inhibition (MacInnes et al., 2003; Tuunanen et al., 2008; Dedkova et al., 2013) that could not completely explain its effectiveness as a cardioprotective agent. However, the fact that these studies used either a non‐optimal model (such as animals or a human atrial cell line, which had lost most of its cardiomyocyte phenotype and is considered a HeLa derivative) or employed indirect methods to obtain fatty acid oxidation measurements resulted in TMZ still being broadly considered to be a metabolic modulator.

The aim of the current study was to obtain a more direct answer to the question if and to what extent TMZ influences cardiac mitochondrial utilization of the main substrates. For this purpose, we measured mitochondrial fatty acid and carbohydrate oxidation in the target tissue of a target patient population, that is, in left ventricular (LV) myocardium which was obtained from CAD patients undergoing surgery for coronary artery bypass grafting (CABG). Two approaches were used: firstly, mitochondrial function and substrate oxidation in myocardial samples taken from patients on prolonged TMZ therapy were compared with those from patients not receiving TMZ (Control). With this approach, any potential chronic effects of TMZ on myocardial bioenergetic remodelling (persisting even after washout of the drug) were tested. Secondly, TMZ was administered in vitro to determine whether the acute administration of TMZ can modify mitochondrial substrate oxidation of human LV myocardium.

Methods

Study design

Forty‐six haemodynamically stable CAD patients scheduled for elective CABG at the University Hospital Split were included in the study. Emergency patients, patients with LV ejection fraction (LVEF) below 30% and patients with concomitant valve replacement and severe renal, hepatic or pulmonary disease were excluded. The patients selected were allocated to one of the two groups: TMZ group [n = 21; Preductal MR, Servier, Neuilly‐sur‐Seine, France; 2 × 35 mg·day⁻¹ – daily recommended dose (Montalescot et al., 2013)] and control group (CTL, n = 25) who had not been exposed to the drug. In the TMZ group, the minimum duration of TMZ therapy prior to surgery was 30 days [chosen based on reports that showed improved cardiac function by TMZ after this time period (Lu et al., 1998)], with median duration of 105 days, and 19 out 21 patients being exposed to TMZ for ≥77 days. According to the ESC guidelines, TMZ is a class IIb drug recommended for treatment of stable CAD, that is, its use as a second‐line angina treatment is optional and left to the cardiologist's judgement (Montalescot et al., 2013). Also, the decision to place each patient on TMZ therapy was made prior to their inclusion in our study, and was made by a cardiologist not involved in this study and was not influenced by the researchers conducting this investigation. In order to compare the patients' characteristics and ensure that there were no major differences between the groups, the clinical data were collected immediately before surgery. In the TMZ group, the drug was withdrawn on the day of surgery (last dose received approximately 12 h before the biopsy). Pre‐, intra‐ and post‐surgical procedures were performed according to the standard clinical routines of the Department of Cardiac Surgery.

The potential effect of chronic TMZ therapy was tested by comparing measurements obtained in TMZ patients versus CTL patients. The effects of acute in vitro administration of TMZ were investigated in tissues obtained from both patient groups (n = 46), where recordings were performed in the presence and absence of the drug. Some measurements, such as positive control and those testing various concentrations of TMZ and substrate, were performed on samples obtained from a subset of six patients who had not received TMZ therapy, had no history of diabetes mellitus (DM) and exhibited preserved systolic function. Data acquisition in experiments testing the chronic effects of TMZ and the analysis for both chronic and acute effects of the drug were performed in a blinded fashion.

The study complies with the Declaration of Helsinki and was approved by the Ethical Committees of the University Hospital Split (2181‐147‐01) and Split School of Medicine (2181‐198‐03‐04). All patients signed an informed consent form before being enrolled in the study. This was a single centre, prospective trial, registered as an observational study at http://www.clinicaltrials.gov under identification number NCT02152527. The experimental design, analysis and presentation comply with the guidance for publication in British Journal of Pharmacology (Curtis et al., 2015).

Left ventricular biopsies

At the end of the CABG procedure, performed without the use of cardiopulmonary bypass and cardioplegia (‘off‐pump’), two to three cylinder‐shaped biopsies (approximately 20 × 1 mm) were taken from the anteroseptal part of the LV, a cardiac region supplied by the left anterior descending (LAD) branch of the left main coronary artery. The biopsy tissue was immediately immersed in cold mitochondria‐preserving storage solution (Solution S, in mmol·L⁻¹: 2.77 CaK₂EGTA, 7.23 K₂EGTA, 6.56 MgCl₂, 5.7 Na₂ATP, 15 phosphocreatine, 20 imidazole, 20 taurine, 0.5 DTT and 50 K‐methanesulfonate, pH 7.1 at 0°C), kept on ice and transferred to the laboratory within 15 min. A portion of the tissue was immediately snap‐frozen for later analyses of enzyme activity and expression, while the remaining tissue was used immediately to obtain the mitochondrial respiration measurements.

Mitochondrial respiration

The myocardial biopsy samples were finely dissected on ice in Solution S under a microscope and then permeabilized by mild agitation at 4°C with the addition of saponin (50 μg·mL⁻¹) (Kuznetsov et al., 2008). Before the respiration measurements were determined, the samples were washed by agitation for 10 min at 4°C in the respiration medium (in mmol·L⁻¹: 2.77 CaK₂EGTA, 7.23 K₂EGTA, 1.38 MgCl₂, 3 K₂HPO₄, 20 imidazole, 20 taurine, 0.5 DTT, 90 K‐methanesulfonate, 10 Na‐ methanesulfonate and 0.2% BSA, pH 7.1; 100 nmol·L⁻¹ free Ca^{2 +}, 1 mmol·L⁻¹ free Mg^{2 +}) and then transferred into a 2 mL respirometry chamber filled with the same solution. Mitochondrial respiration was evaluated using an oxygen Clark‐type electrode (Oxygraph, Hansatech Instruments, Norfolk, UK) at 30°C. Fatty acid oxidation was assessed using palmitoylcarnitine as a substrate (40 μmol·L⁻¹ in the presence of 5 mmol·L⁻¹ malate) and carbohydrate oxidation was assessed by providing mitochondria with pyruvate (10 mmol·L⁻¹) and malate (5 mmol·L⁻¹). Tissue oxygen consumption was recorded in the absence of any substrates, in the presence of substrates only (state 2) and following the addition of saturating amounts of ADP (2.5 mmol·L⁻¹; state 3) and trifluorocarbonylcyanide phenylhydrazone (FCCP, 1 μmol·L⁻¹), a drug that uncouples mitochondrial oxygen consumption from ATP production and thus removes any potential oxidation limitation of the phosphorylation apparatus. Mitochondrial respiratory control ratio, which is an indicator of coupling between phosphorylation and oxygen consumption, was calculated for each metabolic substrate as a proportion of state 3/state 2 respiration. The level of ambient oxygen available was maintained above 210 μmol·L⁻¹ to avoid its diffusion limitation in the fibres (Gnaiger, 2009). Oxygen consumption rate was expressed as pmol O₂·s⁻¹·mg⁻¹ of wet tissue weight.

A subset of LV biopsies was prepared for respiration measurements in saponin‐free conditions using a mechanical homogenization system specifically designed for this purpose (PBI shredder SG3; Pressure BioSciences, South Easton, MA, USA) (Eigentler et al., 2014; Larsen et al., 2014). Cardiac tissue samples were weighed and transferred into pre‐chilled shredder tubes filled with ice‐cold respiration medium. The samples were then homogenized (10 s at medium force and 2 × 5 s at high force) and the volume of homogenate was adjusted with respiration medium for the planned number of measurements and transferred in 2 mL portions into the respirometry chambers. Using this approach, the possibility of the fibre saponin‐permeabilization procedure interfering with the potential effect of TMZ on enzyme activity was circumvented.

Citrate synthase activity assay

Previously snap‐frozen tissue samples were homogenized using a 0.2 mL tissue grinder (Micro tissue grinder; Wheaton, Millville, NJ, USA) in an ice‐cold PBS (1/17 wt per volume) in the presence of a protease inhibitor cocktail (P8340; Sigma‐Aldrich). Protein extraction was performed using 15% lauryl maltoside detergent solution (ab109858; Abcam, Cambridge, UK). Protein concentration was determined with a detergent compatible protein assay (Bio‐Rad, Hercules, CA, USA). The activity of citrate synthase in the reaction mixture containing 15 μg of tissue protein was assessed at 30°C using a kit (CS0720; Sigma‐Aldrich). Enzyme activity was initiated by adding oxaloacetic acid (10 mmol·L⁻¹), and absorbance was measured at 412 nm using a spectrophotometer (DU 800; Beckman Instruments, Fullerton, CA, USA). Upon subtraction of background absorbance, the enzyme activity was calculated using the 13.6 (mmol L⁻¹)^{− 1·}cm⁻¹ extinction coefficient and expressed in international units of citrate synthase activity mg⁻¹ tissue protein (U mg⁻¹).

Pyruvate dehydrogenase assay

Myocardial tissue was homogenized in PBS using a tissue grinder, with added protease inhibitor cocktail (P8340), 20 mmol·L⁻¹ NaF (phosphatase inhibitor) and 1% apyrase (ATP depletion system, 400 U·mL⁻¹) in 1/10 w.v^‐1 ratio. Upon homogenization and protein extraction, pyruvate dehydrogenase (PDH) activity was measured using a PDH enzyme activity kit (AAMT008‐1KIT; Merck‐Millipore, Darmstadt, Germany). The absorbance was measured at 450 nm in samples containing 40 μg of protein, using a microplate reader (EL808 Ultra Microplate Reader; Bio‐Tek Instruments, Winooski, VT, USA) at 30°C. Endogenous PDH enzyme activity was expressed as a rate of absorbance change (mOD·min⁻¹·mg⁻¹ tissue protein) after subtraction of the background signal.

Western blotting

Frozen tissue samples were homogenized using a tissue grinder in modified RIPA buffer supplemented with protease and phosphatase inhibitors. Following SDS‐PAGE and transfer, the nitrocellulose membranes were probed with MitoProfile PHD WB antibody cocktail (ab110416; Abcam, Cambridge, UK) containing four different mouse monoclonal antibodies reacting specifically with E1α, E1β, E2 and E2/E3bp subunits of PDH and total OXPHOS human antibody cocktail (ab110411; MitoSciences, Eugene, OR, USA) containing mouse monoclonal antibodies against structural components of the five mitochondrial respiratory complexes (the NDUFB8 subunit of complex I, the SDHB subunit of complex II, the QCR2 subunit of complex III, the COX II subunit of complex IV and the ATP5A subunit of complex V). After incubation with the corresponding secondary antibodies and SuperSignal West Femto Chemiluminescent Substrate (Pierce/Thermo Fisher Scientific, IL, USA), blots were imaged using Chemidoc imaging system (Bio‐Rad, Hercules, CA, USA). β‐actin served as a loading control. In order to allow for gel‐to‐gel comparison, a standard sample was loaded on each gel, and all tested protein bands were normalized to the standard reference band. All immunoblotting data pertaining to CTL and TMZ groups are expressed as a percentage of the CTL group's average value. The chemiluminescence intensities were measured using image lab 3.0 software (Bio‐Rad).

Statistical analysis

The data and statistical analysis comply with the recommendations on experimental design and analysis in pharmacology (Curtis et al., 2015). We estimated that a sample of 21 patients per group was required for 80% statistical power, based upon previous reports on TMZ's effect on mitochondrial fatty acid oxidation (α = 0.05 and β = 0.2) (Dedkova et al., 2013). Pearson's χ ² test was used for categorical parameters and Student's unpaired or paired t‐tests were conducted to compare between‐group and intra‐group differences in the results. In several experiments where samples from a smaller number of patients were used, non‐parametric statistical tests were employed; Mann–Whitney U‐test for comparing two patient groups and Friedman test (non‐parametric version of the repeated measures anova) for comparison of multiple experimental conditions recorded on biopsy specimens from each patient. In the case of the latter, when the F value was found to be significant, a post hoc Bonferroni test was used to assess the differences between experimental conditions. Statistical analysis was performed using statistica 8.0 software (StatSoft Inc., Tulsa, OK, USA), and a two‐sided P‐value <0.05 was considered significant. Data are presented as means ± SD.

Chemicals

All chemicals used for this study, unless otherwise noted, were purchased from Sigma‐Aldrich (St. Louis, MO, USA). The TMZ used for in vitro administration (freshly prepared) was obtained from Servier.

Results

General observations

At the time of surgery, the two patient groups were assessed as having similar clinical parameters and characteristics (Table 1). All patients successfully underwent the CABG and no complications related to the LV biopsy procedure were evident.

Table 1.

Patient characteristics and perioperative parameters

	Control (n = 25)	TMZ (n = 21)	P‐value
Female gender; n (%)	5 (20)	5 (24)	0.76
Age; mean ± SD	66 ± 8.2	67 ± 6.6	0.73
Euroscore II; mean ± SD	2.71 ± 2.1	2.53 ± 1.7	0.76
LVEF ˃50%; n (%)	22 (88)	18 (86)	0.82
LAD occl >70 %; n (%)	19 (76)	16 (76)	0.74
Number of angina attacks per week; mean ± SD	2.3 ± 0.9	2.1 ± 1.1	0.67
NTG consumption per week; mean ± SD	1.9 ± 0.7	1.8 ± 0.6	0.75
Comorbidity; n (%)
Previous acute coronary syndrome	13 (52)	11 (52)	0.98
Angina pectoris	7 (28)	7 (33)	0.70
Previous myocardial infarction	5 (20)	5 (24)	0.76
Hypertension	16 (64)	12 (57)	0.64
Peripheral arterial disease	1 (4)	4 (19)	0.10
Diabetes Mellitus	6 (24)	7 (33)	0.71
COPD	2 (8)	1 (5)	0.66
Drug therapy; n (%)
Statins	17 (68)	18 (86)	0.16
Calcium antagonist	4 (16)	2 (10)	0.52
β blocker	19 (76)	16 (76)	0.99
ACE‐inhibitor/ATII inhibitor	16 (64)	12 (57)	0.64
Diuretics	11 (44)	6 (29)	0.28
Acetylsalicylic acid	22 (88)	18 (86)	0.82
Clopidogrel	14 (56)	11 (52)	0.81
Organic nitrates	7 (28)	4 (19)	0.48
Metformin	4 (16)	4 (19)	0.91
Sulfonylurea	4 (16)	1 (5)	0.22
Insulin	2 (8)	2 (10)	0.86
LWMH	5 (20)	6 (29)	0.50
Omega‐3 fatty acid	4 (16)	5 (24)	0.51
Biochemical parameters; mean ± SD
Preoperative troponin C(ng·mL−1)	1.16 ± 3.63	0.04 ± 0.05	0.21
Postoperative troponin Ca (ng mL‐1)	3.44 ± 5.65	1.04 ± 0.87	0.09
HbA1C	0.06 ± 0.01	0.09 ± 0.12	0.35
CK‐MB (U·L−1)	9.62 ± 11.12	9.81 ± 4.82	0.95

Values are n (%); mean ± SD. TMZ, trimetazidine‐treated patients; LVEF, left ventricular ejection fraction; LAD occl., % of luminal occlusion of left anterior descendent branch; NTG, nitroglycerin; COPD, chronic obstructive pulmonary disease; AT‐II, angiotensin II receptor; LWMH, low MW heparin; HbA1C, glycated Hb; CK‐MB, MB type of creatine kinase.

^a Measured 24 h after the end of surgery.

Impact of chronic TMZ therapy on myocardial fatty acid and carbohydrate oxidation

The enzymatic activity of citrate synthase, a marker of mitochondrial content in the tissue, was not different between the two groups of patients (Figure 1A). Direct measurement of ADP‐supported respiration in LV myocardium revealed no difference in the rate of palmitoylcarnitine (40 μmol·L⁻¹) oxidation between the CTL and TMZ groups of patients (Figure 1B). The same was observed upon addition of the uncoupler FCCP (Figure 1B). The rate of mitochondrial oxidation of pyruvate, the specific substrate of carbohydrate metabolism, was also not different between the CTL and TMZ groups (Figure 1C), and administration of FCCP increased the rate of oxygen consumption to the same degree in both groups. Also, no difference between the groups was observed with regard to the respiratory control ratio values for palmitoyl and pyruvate‐fuelled respiration (inserts, Figure 1B and C), or in the ratio of carbohydrate to fatty acid substrate oxidation (Figure 1D).

Figure 1.

Mitochondrial fatty acid and pyruvate oxidation in patients receiving trimetazidine (TMZ) therapy and controls (CTL). (A) Activity of citrate synthase, a marker of mitochondrial tissue content. (B) Mitochondrial respiration of tissue alone (tissue), in the presence of palmitoylcarnitine (Pc, 40 μmol·L⁻¹) and with subsequently added saturating amount of ADP (2.5 mmol·L⁻¹) and mitochondrial uncoupling agent FCCP (1 μmol·L⁻¹). (C) Mitochondrial oxidation of pyruvate, a specific substrate of carbohydrate metabolism. No differences were observed between patients receiving chronic TMZ therapy and CTL. Respiratory control ratios (RCR) are shown in inserts. (D) Ratio of mitochondrial carbohydrate oxidation to fatty acid oxidation recorded under same conditions was not different for TMZ and CTL patients. M, malate, n = 21 in TMZ group and 25 in CTL group. *P < 0.05 versus respiration upon addition of substrates for both groups. ww, wet tissue wt.

The proportion of carbohydrate and fatty acid substrate oxidation relative to maximal mitochondrial respiration [Rmax, evoked by addition of both carbohydrate and fatty acid substrates, as well as glutamate (10 mmol·L⁻¹) and succinate (15 mmol·L⁻¹) (Gnaiger, 2009)] is displayed in Figure 2. As seen in the figure, the ADP‐supported oxidation of palmitoyl relative to the Rmax, which was further increased by the addition of FCCP, was not different between the CTL and TMZ groups of patients. The oxidation of pyruvate, relative to the Rmax, was also not different between the two groups (Figure 2B).

Figure 2.

Mitochondrial fatty acid and pyruvate oxidation presented relative to maximal mitochondrial respiration. Maximal rate of oxygen consumption was evoked by addition of substrates for both metabolic pathways, as well as glutamate (G) and succinate (S). Values for fatty acid (A) and pyruvate (B) oxidation were not different between patients on trimetazidine (TMZ) therapy and controls (CTL). Arrows indicate addition of specific substrates during a respiration recording protocol. M, malate, Pc, palmitoylcarnitine, P, pyruvate, n = 21 in TMZ group and 25 in CTL group. *P < 0.05 versus respiration upon addition of substrates for both groups.

Since DM and chronic heart failure have been reported to modify cardiac metabolism (Lopaschuk et al., 2010), we performed a separate analysis of mitochondrial substrate oxidation in non‐DM patients with preserved systolic function (LVEF >50%, n = 9 patients for TMZ and 14 for CTL) and no differences were observed (Figure S1, Supplementary materials).

Effect of chronic TMZ therapy on PDH activity and expression

The proposed effect of chronic TMZ therapy on cardiac metabolism was further investigated through analysis of the activity of PDH, the key enzyme that regulates carbohydrate oxidation in mitochondria. Preservation of the in vivo PDH phosphorylation status (which is the main regulator of the enzyme activity) was ensured by immediate snap freezing of biopsy tissue and subsequent inclusion of phosphorylation and phosphatase inhibitors in the experimental solutions. Our measurements revealed no differences in PDH activity of LV myocardium between CTL and TMZ patients (Figure 3A). Furthermore, measurement of the expression of PDH complex subunits in cardiac tissue also revealed no differences between the two patient groups (Figure 3B and C).

Figure 3.

Impact of (TMZ) therapy on myocardial pyruvate dehydrogenase (PDH) activity and expression. (A) Myocardial activity of PDH, the key regulator of carbohydrate metabolism, is expressed as the rate of change in absorbance at 450 nm (OD) and normalized per tissue mass. (B) Image of the representative SDS‐PAGE blot probed with monoclonal mouse antibody cocktail targeted against E1 (α and β), E2 and E3 subunits of the PDH complex. (C) The chemiluminescence intensity was expressed (after normalization to the loading control) relative to average chemiluminescence intensity of control (CTL) group, which was set to 100% (for each of the subunits). n = 21 in TMZ group and 25 in CTL group.

Acute in vitro TMZ administration and substrate oxidation

The acute in vitro exposure of permeabilized myocardium to TMZ was performed to test whether washout of the drug following cardiac biopsy was the reason for the lack of effect of TMZ therapy on substrate oxidation. For this purpose, the tissue was treated with TMZ (10 μmol·L⁻¹) for 30 min before and throughout the respiration measurements. As presented in Figure 4A and B, no difference in mitochondrial oxidation of any of the substrates added (palmitoylcarnitine 40 μmol·L⁻¹) was observed when TMZ was included in the experimental solutions (n = 46 patients). Conversely, the addition of 4‐pentenoic acid (100 μmol·L⁻¹), an inhibitor of β‐oxidation (Kantor et al., 2000) that was used as a positive control (n = 6 patients), resulted in a reduced rate of palmitoyl oxidation (Figure 4C). Under the same experimental conditions, two additional TMZ concentrations (1 and 100 μmol·L⁻¹) were tested using the cardiac tissue of six non‐DM patients (LVEF >50%) who were not previously exposed to TMZ therapy. Again, in vitro TMZ administration did not affect mitochondrial fatty acid and carbohydrate oxidation (Figure 4D).

Figure 4.

Effect of acute in vitro administration of trimetazidine (TMZ) on myocardial fatty acid oxidation. (A) Mitochondrial palmitoylcarnitine (Pc, 40 μmol·L⁻¹) oxidation recorded upon addition of ADP and FCCP was not different, irrespective of the presence of TMZ (10 μmol·L⁻¹) in the experimental solutions. Subsequent additions of other substrates resulted in the same degree of oxidation increments (n = 46). (B) Individual respiratory states expressed relative to maximal mitochondrial respiration. In two separate subsets of samples (n = 6 patients each), fatty acid oxidation was also recorded in the presence of β‐oxidation inhibitor 4‐pentenoic acid [(C) first subset] and with two additional TMZ concentrations [(D) second subset]. Pc40, palmitoylcarnitine 40 μmol·L⁻¹, M, malate; P, pyruvate; G, glutamate; S, succinate. *P < 0.05 versus respiration upon addition of substrates for both experimental groups. **P < 0.05 versus no TMZ and TMZ in vitro. ww, wet tissue wt.

To control for the possibility that saponin, used for permeabilization of myocardial fibres, interfered with any potential effects of TMZ on enzyme activity, palmitoylcarnitine oxidation was also recorded in myocardial tissue homogenates (without saponin) that were obtained from a subset of non‐DM patients with preserved systolic function (n = 6, not previously exposed to TMZ). Using this approach, we detected no effect of TMZ (applied at concentrations of 1, 10 and 100 μmol·L⁻¹ in vitro) on mitochondrial fatty acid oxidation (Figure S2, Supplementary materials).

Moreover, because it has been reported that the ability of TMZ to suppress β‐oxidation depends on the concentration of fatty acid substrate available (Lopaschuk et al., 2003), we also measured mitochondrial respiration using several lower palmitoylcarnitine concentrations (ranging from 1 to 15 μmol·L⁻¹) in samples from non‐DM patients with preserved systolic function (n = 6 patients, not previously exposed to TMZ). As seen in Figure 5, mitochondrial palmitoyl oxidation was not affected by the in vitro administration of TMZ (10 μmol·L⁻¹) at any of the palmitoylcarnitine concentrations tested [4‐pentenoic acid (100 μmol·L⁻¹) was used as a positive control].

Figure 5.

Probing the effect of trimetazidine (TMZ) administration in vitro on cardiac fatty acid oxidation using lower concentrations of palmitoylcarnitine (1–15 μmol·L⁻¹). Mitochondrial fatty acid oxidation was assessed in the presence of ADP and FCCP. No effect of acute TMZ administration (10 μmol·L⁻¹) with any of the tested palmitoylcarnitine (Pc) concentrations was observed. M, malate, n = 6 patients. Stepwise increments of Pc concentration resulted in a significant increase in oxygen consumption within the same experimental group (P < 0.05), *P < 0.05 versus no TMZ and TMZ in vitro. ww, wet tissue wt.

Discussion

The main finding of our study is that TMZ does not alter the oxidation of fatty acids and carbohydrates in LV myocardium of patients suffering from CAD and undergoing CABG. By comparing mitochondrial substrate utilization in myocardium of CTL and TMZ patients, no effects of chronic TMZ therapy were found, suggesting that TMZ does not induce the remodelling of myocardial bioenergetic machinery. Moreover, by acutely exposing the permeabilized myocardium to TMZ in vitro, the effects of the drug washout were avoided; however, TMZ again did not influence mitochondrial substrate oxidation. These findings contradict the generally accepted mechanism of TMZ's mode of action, whereby its beneficial clinical effects are attributed to the cardiac metabolic switch towards more efficient utilization of glucose. However, this widespread belief on TMZ's mechanism of action was formed based on either indirect or animal studies (Kantor et al., 2000; Lopaschuk et al., 2003). The significant advantage of the current study is that the hypothesized metabolic effects of TMZ were measured in tissue biopsies taken from the LVs of CAD patients, enabling direct measurements of mitochondrial substrate oxidation in a target tissue of the patient population in which TMZ therapy is primarily indicated.

According to Randle (Randle et al., 1963), partial suppression of fatty acid oxidation results in the compensatory augmentation of the carbohydrate metabolic pathway, which is the primary mechanism proposed to underlie the beneficial clinical effects of TMZ in the ischaemic heart (Kantor et al., 2000; Lopaschuk et al., 2010). By comparing the myocardial substrate oxidation of CAD patients chronically treated with TMZ with patients who were not exposed to the drug, we aimed to investigate whether this metabolic reprogramming can be induced by TMZ therapy. In this case, TMZ‐induced effects would most likely be present, regardless of the drug washout from the time of biopsy until respirometry measurements. Our experiments revealed that neither oxidation of fatty acid (palmitoylcarnitine) nor carbohydrate (pyruvate) substrates was affected by the long‐term TMZ treatment. In addition, the endogenous activity and expression of PDH, the key regulator enzyme of carbohydrate metabolism, were not altered by the chronic TMZ therapy. This finding further supports our respirometry data, because several studies have indicated that, if glucose uptake and oxidation is increased, PDH activity should be upregulated as well (Lauritzen et al., 2013). Our results differ from the previously reported increase in the active form of PDH found in isolated rat hearts perfused with TMZ (Kantor et al., 2000). Conversely, pyruvate oxidation was not increased in rat cardiac mitochondria isolated from TMZ‐treated rats or after in vitro administration of TMZ, despite a concomitant reduction in palmitoylcarnitine oxidation (Fantini et al., 1994).

In our experiments, acute in vitro administration of TMZ [10 μmol·L⁻¹, a concentration was based on previous reports (Kantor et al., 2000; MacInnes et al., 2003)] also failed to inhibit mitochondrial β‐oxidation. Similar results were obtained with two additional TMZ concentrations, tested in a subset of samples (1 and 100 μmol·L⁻¹). This finding contrasts with previous reports showing that acute TMZ administration suppresses mitochondrial fatty acid fuelled respiration. For example, in cardiac mitochondria isolated from Wistar rats, the inclusion of TMZ in the respiratory medium decreased the oxidation of palmitoylcarnitine by ~40% (Fantini et al., 1994). In a study by Dedkova (Dedkova et al., 2013), although TMZ‐induced cardioprotection was primarily ascribed to non‐metabolic effects of the drug, the authors still detected a somewhat reduced fatty acid oxidation (~15%) in permeabilized rabbit cardiomyocytes exposed to 1 μmol·L⁻¹ TMZ in vitro. The experimental approach that we used to assess mitochondrial substrate oxidation, high‐resolution respirometry in saponin‐permeabilized myocardial fibres, is known to have many advantages over an analysis in isolated mitochondria, as it enables better preservation of an integrated cellular system (Kuznetsov et al., 2008). Some of the potential methodological drawbacks, such as heterogeneity of mitochondrial density in biopsy samples, were eliminated by normalizing the oxidation rates with reference to maximal mitochondrial respiration (Lemieux et al., 2011). Moreover, the activity of the mitochondrial marker enzyme, citrate synthase, assessed in regions from which the biopsy samples were taken, was not different between the CTL and TMZ groups. A set of respiratory measurements was also performed in LV tissue homogenates (rather than the saponin‐permeabilized fibres) and confirmed that saponin was not the cause for the lack of effect of TMZ. Finally, the ability of our experimental set‐up to detect small, yet significant changes in fatty acid oxidation, such as during a gradual increase in palmitoyl concentration and upon addition of a positive control, validated its sensitivity for detecting the potential effects of TMZ in the order of magnitude as those reported previously (Fantini et al., 1994; Dedkova et al., 2013).

Another major difference from previous studies is that our measurements were performed in human LV tissue, and species differences may provide an explanation for the contrasting results. Indeed, MacInnes et al. (2003) also did not observe any effect of acute TMZ administration on β‐oxidation in a cultured human cell line (Girardi cells). These data were complemented with direct measurements of the enzymatic activity of LC 3‐KAT, the suggested target enzyme of TMZ's action, from recombinant human trifunctional protein. Again, a range of TMZ concentrations was observed to have no inhibitory effects. The main concern raised about the MacInnes' study was that an unphysiologically high concentration of 3‐keto‐hexadecanoyl CoA, a substrate for LC 3‐KAT was used (Lopaschuk et al., 2003). In their subsequent work, Lopaschuk et al. have shown, using an animal model, that a substrate concentration in excess of 15 μmol·L⁻¹ can overcome the inhibitory effect of TMZ on β‐oxidation, suggesting that TMZ's action is mediated via competitive blockade of LC 3‐KAT. When lower concentrations of substrate were used (2.5–10 μmol·L⁻¹), the authors recorded a reduced activity of LC 3‐KAT in the presence of TMZ. For this reason, we performed an additional set of experiments, in which the LV fibres were subjected to a range of lower palmitoylcarnitine concentrations (1–15 μmol·L⁻¹). However, we still did not observe any effect of in vitro TMZ administration on mitochondrial fatty acid oxidation.

It should be emphasized that the purpose of our investigation was not to validate or re‐evaluate the clinical efficacy of TMZ, as our study did not include a prospective monitoring of cardiac function (e.g. LVEF) before and after TMZ therapy. The cardioprotective potential of the drug has been reported in a number of other studies (Szwed et al., 2001; Fragasso et al., 2013; Vitale et al., 2013). Our data on the lack of effect of TMZ on mitochondrial substrate oxidation are also indirectly supported by results of some of the trials dealing with the clinical efficacy of TMZ. For example, TMZ therapy was found to be beneficial in some cardiac diseases where the main underlying pathophysiology does not involve ischaemia and oxygen limitation [the main disturbances that TMZ is proposed to ameliorate through the suggested increase in efficiency of ATP production and an improved coupling of glycolysis with glucose oxidation (Onbasili et al., 2007; Aygen et al., 2008; Ferraro et al., 2013)]. Along this line, an improved cardiac function following 3 months of TMZ therapy was also reported recently in patients suffering from idiopathic dilatative cardiomyopathy (Tuunanen et al., 2008). Using PET scanning, the authors detected only a modest (10%) reduction in myocardial free fatty acid uptake and concluded that the energetic improvements potentially resulting from this effect are the unlikely mechanism of TMZ's positive action.

Interestingly, another anti‐anginal agent from the group of metabolic modulators, ranolazine, was also, as TMZ, assumed to act via β‐oxidation inhibition (McCormack et al., 1996). However, recent studies have shown that ranolazine in fact inhibits the late Na⁺ inward current (Sossalla et al., 2008), which provides a more plausible mechanism for its beneficial action (prevention of cardiac myocytes Ca^{2 +} overload occurring secondary to cytosolic Na⁺ increase) (Hale et al., 2006; Sossalla et al., 2008). Furthermore, a potential side effect of therapies aimed at inhibiting cardiac fatty acid oxidation might be an enhanced accumulation of lipids in the cytosol (Sharma et al., 2004), which was reported to induce pathological remodelling (Wende and Abel, 2010) and is likely to counteract any beneficial effects these drugs might exert on cardiac energetic efficacy. Alternatively, TMZ action on several pathways other than β‐oxidation has recently been suggested. In a study on a rabbit heart failure model, the protective action of TMZ observed in isolated cardiac myocytes was primarily linked to a reduction in mitochondrial permeability transition pore (mPTP) opening and diminished generation of ROS via modulation of complex II activity and mitochondrial NOS, while administration of another 3‐KAT inhibitor, 4‐bromotiglic acid, was not cardioprotective (Dedkova et al., 2013). Suppressed ROS generation was also detected in studies by Iskesen et al. (2006), Liu et al. (2010) and Dehina et al. (2013), where this effect was observed after only 4 days of oral TMZ treatment in a pig acute ischaemia model, along with preserved mitochondrial structure and cardiomyocyte protection. The possible mechanism of TMZ including effects on mPTP was also explored by Argaud et al., 2005, who reported that an infusion of TMZ prior to ischaemia/reperfusion in NZW rabbits prevented mPTP opening. Moreover, in a study by Tritto et al. (2005), TMZ‐induced protection of contractile function in isolated rat heart during ischaemia/reperfusion was linked to inhibition of neutrophil activation, while Chen et al. (2015) detected inhibition of the macrophage‐mediated pro‐inflammatory response in mice treated with TMZ prior to endotoxin‐induced cardiomyopathy. Therefore, TMZ might also act via some extramitochondrial, or even extracardiac mechanisms, as suggested by recent studies reporting the effect of TMZ on endothelial function (Rehberger‐Likozar and Sebestjen, 2015) and whole body insulin sensitivity (Tuunanen et al., 2008). Some of these effects may, independently of metabolic modulation, confer the clinical improvements associated with TMZ therapy.

Study limitations

There are several points related to the characteristics of our study that need to be considered. CAD patients are still a heterogeneous population with differences in medications and co‐morbidities. However, a comparison of the various clinical and biochemical parameters between the two patient groups (as seen in Table 1) shows that they were well matched. Also, some other medical conditions, such as DM and heart failure, have been reported to interfere with cardiac metabolism (Lopaschuk et al., 2010; Nagoshi et al., 2011). To exclude the possibility that the potential effect of TMZ therapy is masked by the metabolic modulation elicited by these conditions in both patient groups, we performed a separate analysis excluding all the patients with DM and systolic dysfunction. Furthermore, we evaluated myocardial fatty acid and carbohydrate oxidation through direct measurements of oxygen consumption by mitochondria provided with substrates specific for either metabolic pathway. However, deficiencies in several mitochondrial elements, such as the phosphorylation apparatus (Lemieux et al., 2011), have been reported in chronically ischaemic heart, which could potentially affect our results. For example, a reduced action of ATP synthase or the adenine nucleotide translocator might slow the electron flow down the electron transfer chain (ETC), reduce the rate of oxygen consumption and eliminate the potential effects of TMZ on substrate metabolic turnover occurring upstream. For this reason, we also employed substances such as FCCP, which uncouple the mitochondrial oxygen consumption from the ATP production and remove the potential inhibitory effect of the phosphorylation apparatus. Also, differences in expression of ETC proteins could mask the respiration results. To avoid this possibility, we performed Western blot analysis of representative subunits for all five ETC complexes, which revealed no differences between the two groups (Figure S3, Supplementary materials). Furthermore, a question can be raised regarding the washout of TMZ from the tissue of TMZ patients and how this affects the interpretation of the results pertaining to chronic TMZ therapy. However, there are many examples of lasting cardiac effects of drug exposure that are present after drug washout – for instance, cardioprotection in pharmacological preconditioning (e.g. by volatile anaesthetics), which lasts long after drug elimination (first and even second window of preconditioning appearing after approximately 24 h) (Feng et al., 2008). Moreover, with experiments in which the acute exposure to TMZ was tested, we also addressed this issue. Lastly, it should be emphasized that the myocardial biopsies were performed during an ‘off‐pump’ surgery, meaning that the potential metabolic effects of cardiac ischaemia and various cardioplegia interventions that are present during an ‘on‐pump’ surgery were avoided.

Conclusion

In summary, we found no evidence of either a chronic or acute effect of TMZ on cardiac fatty acid and carbohydrate oxidation, which is considered a primary mechanism underlying its efficacy as an anti‐anginal medication. Although there are an increasing number of clinical studies showing its clinically advantageous effects, both in CAD and in other ischaemia‐unrelated conditions, our study points towards the need for further translational research aimed at identifying the actual intracellular mechanism responsible for TMZ's cardioprotective actions.

Author contributions

M.C., M.L. and J.M. designed the study, performed the research, analysed the data and wrote the paper; C.B., D.B., D.F., J.K. and N.K. designed the study, performed the research and revised the manuscript; Z.D., C.J.L. and U.W. analysed the data and revised the manuscript.

Conflict of interest

The authors declare no conflicts of interest.

Declaration of transparency and scientific rigour

This Declaration acknowledges that this paper adheres to the principles for transparent reporting and scientific rigour of preclinical research recommended by funding agencies, publishers and other organizations engaged with supporting research.

Supporting information

Figure S1 Mitochondrial substrate oxidation in non‐diabetic patients with preserved systolic function. To exclude the potential effect of heart failure and diabetes on data interpretation, additional analysis was performed in a subgroup of patients with cardiac ejection fraction above 50% and no history of diabetes. A. Mitochondrial respiration of tissue alone (tissue), in the presence of palmitoylcarnitine (Pc, 40 μmol·L⁻¹) and with subsequently added ADP (2.5 mmol·L⁻¹) and FCCP (1 μmol·L⁻¹). B. Mitochondrial oxidation of pyruvate (P). M, malate; G, glutamate; S, succinate; n = 9 in TMZ group and 14 in CTL group. *P < 0.05 versus respiration upon addition of substrates for both groups.

Figure S2 Mitochondrial fatty acid oxidation recorded in cardiac homogenates. The potential confounding effect of saponin permeabilization was eliminated by measuring the palmitoylcarnitine oxidation (Pc, 40 μmol·L⁻¹) in homogenates prepared from myocardial tissue biopsies. Mitochondrial respiration rates recorded upon addition of ADP and FCCP were not affected by the addition of trimetazidine at any of the tested concentrations. M, malate; n = 6 patients.

Figure S3 Expression of mitochondrial electron transfer chain (ETC) respiratory complexes in TMZ‐treated (TMZ) and control (CTL) patients. A. Image of representative Western blot showing bands corresponding to specific subunits of mitochondrial respiratory complexes I, II, III, IV and V. B. Average chemiluminescence intensity normalized to the loading control and expressed relative to average of CTL group, which was set to 100% (for each of the subunits). Chronic trimetazidine therapy did not alter expression of the ETC protein subunits. n = 21 in TMZ group and 25 in CTL group.

Supporting info item

Supporting info item

Supporting info item

Supporting info item

Cavar, M. , Ljubkovic, M. , Bulat, C. , Bakovic, D. , Fabijanic, D. , Kraljevic, J. , Karanovic, N. , Dujic, Z. , Lavie, C. , Wisloff, U. , and Marinovic, J. (2016) Trimetazidine does not alter metabolic substrate oxidation in cardiac mitochondria of target patient population. British Journal of Pharmacology, 173: 1529–1540. doi: 10.1111/bph.13454.