Trimetaridine Effects on the Damage to MiWhondrial Functions Caused by lschemia

and Reperkion Keith Veitch, PhD, Liliane Maisin, and Louis Hue, MD, PhD

Trimetaxidine (TMZ) is an anti-ischemic compound whose precise mode of action is unknown, although seveml studies have suggested a metabolic effect, and there have been reports of protection of mito- chondria against oxidative stress damage. Using a Langendorff rat heart model, we examined the effects of TMZ on the mitochondrial damage follow- ing 30 minutes of ischemia and 5 minutes of reper- fusion. Mitochondrial respiration with succinate, glutamate-malate and ascorbate-N,N,N’,N’-tetm- methylphenylenediamine (TMPD) as substrates was significantly decreased following ischemic+reperfu- sion. Preperfusion with 1 Oe5 M TMZ had no effect on these rates in nonnoxic or ischemic hearts. How- ever, 1 Ow3 M TMZ significantly decreased the gluta- matgmalate rate in mitochondria from normoxic hearts, and this rate was not further decreased following ischemia+-eperfusion, and 1 Oe3 M TM2 also partially protected ascorbat*TMPD activity.

The effect on glutamate-malate was probably due to an inhibition of complex I by TMZ, which specifi- cally inhibited reduced nicotinamide-adenine-di- nucleotide-cytochrome c reductase and complex I in lysed mitochondria. We also studied the effects of TMZ on the activity of pyrvvate dehydrogenase (PDH) in normoxic and ischemic hearts perfused with 0.5 m/W palmitate, which caused the enzyme to be almost completely inactivated. After short peri- ods of ischemia (1 O-20 minutes) the PDH inactiva- tion by palmitate was progressively lost. Preperfu- sion with 1O-5 M TMZ had a tendency to decrease lactate dehydmgenase release, accompanied by a maintenance of the inhibition of PDH by palmitate. This may allow the heart to oxidize fatty acids preferentially during reperfusion, hence removing possible toxic acyl esters.

(Am J Cardioll995; 76:25B-30B)

T rimetazidine (TMZ) is currently used in the treatment of disorders with ischemic origins in cardiology and otolaryngology. The pre-

cise mode of action of the compound is unknown, but experimental studies suggest a metabolic ef- fect. In hypoxic or ischemic hearts TMZ prcservcs the cellular adenosine triphosphate (ATP) levels, decreasing intracellular inorganic phosphate accu- mulation and acidosis during the ischemic cpi- sode.iJ In monocrotaline-induced cardiac hypertro- phy, TMZ decreases the oxidativc damage to mitochondria’ and protects such hearts against the ischemia-reperfusion damage to mitochondrial res- pirati0n.j We have previously shown that the isch- cmia-rcperfusion damage to the mitochondrial respiratory chain is primarily located at the site of complex I in perfused rat hearts.5 In view of the above reports on TMZ, which implicate an effect on mitochondrial metabolism, we applied TMZ to our system to dctcrmine whether the compound could protect against mitochondrial damage dur- ing ischcmia-rcpcrfusion. Subsequently, WC have

From the Hormone and Metobollc Research Unit, lnternotionol

Institute of Cellular and Molecular Pathology, ond University of

LouvaIn Medical School, Brussels, Belgium. This work was supported

by lnstltut de Recherches lnternatlonales Sewer

Address for reprints. Keith Veltch, PhD, HORM unit, ICP-UCL

7529, Avenue Hippocrate, 75, B- 1200 Brussels, Belgium.

studied the influence of TMZ on the activity of pyruvatc dehydrogenase in rat hearts perfused under a variety of conditions. This enzyme plays a key role in switching the cellular metabolism be- tween glycolysis and p oxidation as sources of ATP, and as such may be involved in the selection of the most appropriate energy fuels during ischemia.

METHODS Rat hearts were perfused in a normothermic

Langendorfiapparatus at constant pressure (60 cm H?O) with a modified Krebs-Hens&it buffer in equilibrium with 95% 02/5% CO,.> All hearts were initially perfused for a Bminute equilibration period without recirculation, then subjected to the specific perfusion protocols described. At the end of these treatments, the hearts were either frozen between aluminum plates prccooled in liquid nitro- gen for mctabolite and enzyme assays, or a mito- chondrial fraction was prepared by proteolytic digestion and differential centrifugation,’ which was immediately used to determine respiratory function.

RESULTS AND DISCUSSION Mitochondrial respiration: As already men-

tioned, TMZ has been shown to improve postisch- cmic ATP levels.’ We therefore studied the effects

A SYMPOSIUM: MANAGEMENT OF MYOCARDIAL ISCHEMIA 258

600

500

400

300

200

loo

0

400

300

200

loo

0

1000

Boo

600

400

200

0

A,

B.

FIGURE 1. State 3 respiration rates in mitghondria from nontoxic (open bars) and ischemic+qwfWed (hatched bad hearts with IAl 10 mM succinate I + 2.5 IA ratetwnel; (B) 5’9 glutam&bnd 5 mM bate and 0.25 mM amine as substmtes. Hearts were perfwed for 15 minutes withO,lOpM,orl mMtrimetazidine(TMZ)befareisch- emia, or for 45 minutes of normaxia. TMZ-treated animals receiwdl mg/kgTMZarallytwicedailyfar5daysbefore use,thesehemtsalsabeing withlOpMTMZ.Val- uesaremeans*SEfor5-7 tts per group. Siinificant diirences (p <0.05) are shun versus normoxic cati, ynommxia at the same TMZ, and cc~ntrol ischemh

of TMZ in our system on ATP levels in hearts reperfused for 5 minutes following 30 minutes of global ischemia. This length of ischemia was cho- sen because we have previously shown that shorter periods of ischemia do not engender any decrease

in respiratory function in rat heart mitochondria.5 TMZ was administered from the beginning of the equilibration period, i.e., 15 minutes preperfusion before the ischemia, and was also present in the reperfusion buffer. Ischemia caused a marked decrease in cellular ATP (from 3.31 f 0.39 to 0.62 2 0.10 umol/g wet weight; p < 0.001). Preper- fusion with TMZ had no effect at 10 t&f, but there were significant improvements at 100 m and 1 mit! TMZ (to 1.43 it 0.13 and 1.89 + 0.33 umoI/g wet weight, respectively).

Thus, to study the effects of TMZ on the mitochondrial respiratory function, we chose 10 u.iW and 1 m/t4 TMZ to represent doses that had no effect and caused an improvement of ATP levels, respectively. TMZ was administered again for 15 minutes prior to ischemia and during reperfusion. One group of animals was pretreated with TMZ (1 mg/kg orally twice daily for 5 days), these hearts also being perfused with 10 uJ4 TMZ throughout the equilibration and reperfusion periods. Figure 1 shows the effects of the ischemia and reperfusion on coupled mitochondrial state 3 rates of respira- tion with 3 different sets of substrates that were used to measure activities through different sec- tions of the electron transfer chain (Figure 2). Hence, rates with 10 mM succinate, in the presence of 2.5 uJ4 rotenone, represent the activity through complexes II, III, and IV; 5 mM glutamate with 5 mM malate assays the activity through complexes I, III, and IV; and 5 mM ascorbate plus 0.25 mM N,N,N’,N’-tetramethylphenylenediamine (TMPD) provide an index of the activity of complex IV. As shown, the rate with succinate was significantly decreased by 38% after ischemia-reperfusion (from 536 & 24 to 332 + 24 ng atoms 0 * min-’ * min-I; p <O.OOl), but these rates were not affected by either concentration of TMZ nor by pretreatment with the drug, in either normoxia or ischemia- reperfusion. The rate obtained with glutamate- malate was significantly decreased following isch- emia-reperfusion (from 345 2 12 to 217 + 17 ng atoms 0 * mini * min-*), and this decrease was not affected by 10 t&f TMZ, with or without pretreatment of the animals. However, in normoxic hearts perfused with 1 n&f TMZ the glutamate- malate rate was significantly inhibited by some 16% (289 + 21; p <0.05), and this rate was not further affected by ischemia-reperfusion (271 ? 25). It is debatable whether this represents a protection of the mitochondria, but the effect in normoxic hearts does indicate that TMZ can influ- ence the mitochondrial respiratory chain when administered in the perfusate. Finally, the respira-

268 THE AMERICAN JOURNAL OF CARDIOLOGY VOLUME 76 AUGUST 24, 1995

FIGURE 2. Schema of rates of res- pitution through 3 different sub- strates. cyto = cytochrome c; NADH = reduced nicotinamide- adenine dinudeotide; TMPD = N,N,N’,N’-tetmmethyf- phenylenediamine; TMZ = trimetazidine; UQ = ubiquinone.

Glutamatelmalate Succinate Ascorbate

0

IV

H20

I

tion rate with ascorbate-TMPD was also de- creased following &hernia-reperfusion (from 837 ? 44 to 646 + 37; p <O.OOl), which was not altered by 10 uM TMZ (with and without pretrcat- ment), but there was a small improvement in the presence of 1 mM TMZ (747 + 212; p <0.05 vs ischemia-repcrfusion alone).

These effects on state 3 rates were accompanied by increases in state 4 rates, indicative of small uncoupling effects of the ischemia-reperfusion, which were reflected in markedly lower respiratory control ratios, as well as small but significant decreases in the adenosine diphosphate (ADP) to oxygen ratios (not shown). However, TMZ admin- istration had no effect on any of these other parameters, which was not due to the effects on the state 3 rates.

These changes in coupled state 3 rates were qualitatively the same when measured in un- coupled conditions (1 t&f carbonyl cyanide p-trifluoromethoxyphcnyl hydrazone), i.e., signifi- cant decreases due to ischemia-reperfusion that were not influenced by the presence of TMZ. This shows that they were not due to effects on the ATPase or ATP/ADP translocase activities, but rather they appeared to be caused by alterations of the respiratory chain itself.

Thus, TMZ does not appear to exert any protec- tive action on the damage induced to the mitochon- drial respiratory chain by ischcmia and/or reperfu- sion, even when administered at concentrations that wcrc lOO-l,OOO-fold greater than the effective therapeutic plasma concentration of about 0.15 FM.~ However, the relatively high concentration of 1 mM TMZ did cause a small, but significant, inhibition of glutamate-malate respiration in nor- moxie hearts. As this cffcct was not observed against the other substrates, it was likely to have been caused by a specific inhibition of complex I.

Further, the addition of 1 mMTMZ to the assay cuvcttc had no effect on the respiration rate with any substrate tested in intact mitochondria. How- cvcr, when mitochondria were lysed by repeated

freeze-thawing to expose the reduced nicotinamide- adenine-dinucleotide (NADH) site on the inner face of the inner mitochondrial membrane, TMZ caused a dose-dependent inhibition of NADH- dependent oxygen uptake (Ki = 100 PM; not shown). This was further examined by assaying the NADH-cytochrome c and succinate+ytochrome c reductase activities in control mitochondria (Fig- ure 3), which revealed a specific inhibition of the NADH-cytochrome c reductase (Ki = 100 @4), with no effect on succinate-supported cytochrome

log (Trlmetazldine] (M)

OL Control -6 -5 -4 -3

log [Trimetazidlna] (M)

FIGURE 3. A, the effect of TM! on cytodwome c reduction was tested in rat heart mitochondti lysed by 3 freez+thaw cycles with 100 @I NADH (e) or 25 rnM succinate (0) as sub- strates. B, complex I activity was assayed as the NADH-ubi- quinone reductase activity with 100 mA4 ubiquinone-1 and 130 rJH NADH OS acceptor and substrate, respectively.

A SYMPOSIUM: MANAGEMENT OF MYOCARDIAL ISCHEMIA 278

Regulation of Pyruvate dehydrogenase activity

PDh Active

PDh(P) inactive

FIGURE 4. Schema for the regula- tion of pyrwate dehydragenase (PDh) activity. ADP = adenosine diphosphate; ATP = odenosine triphosphnte; CoA = coenqme A; DCA = dichloroowtab; NAD+ = oxidiied nicotinamide- ode&w dinucleotide; NADH = reduced NAD; Pi = inorganic phosphate.

c reduction at up to 1 mM. The specific inhibition of complex I was confirmed by measuring NADH- ubiquinone-1 reductase, which was inhibited by TMZ (Figure 3). The apparent differences in sensitivity to TMZ in the reductions of cytochrome c and ubiquinone-1 were probably a reflection of the IO-fold difference in protein concentrations required to perform these assays (2 and 20 kg/mL, respectively). This inhibitory effect on complex I is consistent with that of other piperazine derivatives that we have studied,” although TMZ is less potent than cinnarizine and flunarizine (Ki = 10 @4 for both). Although the inhibition of complex I by TMZ is unlikely to be involved in the beneficial effects of TMZ treatment, it may well prove to be the reason why high concentrations of TMZ have been reported to decrease cardiac work’ (600 @!) and to exacerbate ischemic damage’ (100 @4) in perfused rat hearts.

The lack of effect on intact mitochondria in vitro indicates that TMZ acts at a site on the inner face of the inner mitochondriai membrane, which is not accessible in intact organeiies but is exposed when the mitochondria are damaged, e.g., by freeze- thawing. However, the inhibition observed in mito- chondria from perfused hearts suggests that in such conditions there is some uptake of TMZ into the mitochondriai matrix, where it remains through- out the preparative procedure.

Pyruvate dehydrogenase: As we have already indicated, pyruvate dehydrogenase plays a key role in determining whether the mitochondria metabo- lize glucose or lipid fuels for ATP generation. Increased giycoiytic flux and treatments that main- tain a high pyruvate dehydrogenase activity-e.g., activation by dichioroacetate-have been shown to be beneficial during ischemia. It has been sug- gested that it is to the myocyte’s advantage to use giucidic substrates during ischemia-reperfusion, as this is more efficient in terms of ATP provided per

oxygen consumed. Various factors are involved in the interconversion of the active (nonphosphory- iated) and inactive (phosphoryiated) forms, which is controlled by 2 enzymes, pyruvate dehydroge- nase kinase and pyruvate dehydrogenase phospha- tase (Figure 4). These factors include the redox state of the ceil, i.e., the NADH/NAD+ ratio, the oxidative fuels available (ace@-coenzyme A and pyruvate) and the ionic balance of the ceil (Ca2+, Mg2’).

We examined the effects of ischemia-reperfu- sion on pyruvate dehydrogenase activity in our Langendorff rat heart model to determine whether TMZ influences this activity. We routinely assayed the active form of pyruvate dehydrogenase in extracts of freeze-clamped hearts and expressed this activity as a percentage of the total activity, which was determined following the addition of exogenous pyruvate dehydrogenase phosphatase.*

In initial experiments hearts from fed rats (which routinely had a pyruvate dehydrogenase activation state of approximately 30%) were perfused without recirculation for up to 45 minutes. This led to a gradual activation of pyruvate dehydrogenase to about 83%, which is consistent with 11 mM glucose being provided as the sole energy substrate. Fur- ther, when the redox state was altered by inclusion of 10 r&f lactate/l mM pyruvate, there was a marked decrease in the activation to 42% (p <O.Ol), showing that the pyruvate dehydroge- nase system was responsive to such manipulation. However, in order to investigate the effects of lipids on this activity, it was necessary to include 0.5 mM paimitate, bound to deiipidated bovine serum albumin (0.5% wt/voi), in the perfusion buffer. To overcome technical problems, such as oxygenating the buffer without foaming, the heart was perfused in a recirculating mode in a fixed volume of 100 mL. Perfusing in recirculating mode for 30 minutes in the absence of paimitate/aibumin caused the

288 THE AMERICAN JOURNAL OF CARDIOLOGY VOLUME 76 AUGUST 24, 1995

pyruvate dehydrogenasc activation to remain at about 35%. Addition of albumin alone did not affect this value (37%) but the presence of palmi- tate decreased the activation to such an extent that in some hearts it was not possible to detect active pyruvate dehydrogenase before the addition of exogenous pyruvate dehydrogcnase phosphatasc. It was against this background of negligible pyru- vate dehydrogenase activity (2-3%) that we tested the effects of ischemia and TMZ.

Hearts were first equilibrated for 15 minutes in a nonrecirculating system with 11 mM glucose alone, then equilibrated for 15 minutes in the closed circuit system with glucose plus palmitate/ albumin, and ischcmia was then initiated and followed by 5 minutes of reperfusion. TMZ was present during both equilibration periods and in the reperfusion buffer. Samples of perfusate were taken immediately before ischemia, and at the end of the reperfusion period to assay lactate dehydro- genase release as an index of cellular damage, and lactate release as a measure of anaerobic glycolysis. Short periods of ischcmia (10-20 minutes) were used so that no damage would occur to the respira- tory chain.

The effects of different periods of &hernia on the lactate dehydrogenasc release and pyruvate dehydrogenase activation arc shown in Figure 5. There was an appreciable release of lactate dehy- drogenase after only 10 minutes of ischemia, which increased with increasing ischemia. TMZ 10 ~J’V had a tendency to reduce this release, although the effect was only significant after 10 minutes. After 15 minutes of ischemia, 100 ~J’V TMZ also signifi- cantly reduced lactate dehydrogenasc release, dem- onstrating a dose-dependent effect. Lactate re- lease followed a similar pattern to lactate dehydrogenase, increasing according to the dura- tion of the ischemia, but this was not affected by TMZ.

Examination of the pyruvate dehydrogenase activity revealed 2 effects. Firstly, in control hearts, the apparent inhibition of pyruvatc dehydrogenasc by palmitatc was relieved following ischemia in a time-dependent manner, such that after 20 min- utes of ischemia 81% of the pyruvatc dchydroge- nase was active. This suggests some disruption of the control mechanisms by which palmitate nor- mally causes the inactivation of pyruvate dehydro- genase. It is not likely to have been caused by consumption of palmitatc, as the effects of palmi- tatc in normoxia, in which the rate of p oxidation would bc greatest, were present after 30 minutes of perfusion. We are currently investigating the na-

r A.

80

60

40

20

0 10 min 15 min 20 min

FIGURE 5. A, lactate dehydrogenase (LDh) was measured in the perfuwte after 5 minutes of reperfusion following 1 O-20 minutes of ixhemia in control hearts (open ban), or in heartsperfusedwith10~(hatchedbars)or1 mMT?vU (filled bats). B, pytwate dehydtqenose (PDh) was ossayed in extmcts of freeze-clomped hearts. All values are mean f SE of 4-6 hearts.

ture of this apparent change in the pyruvate dehydrogenase regulatory mechanisms.

The second point was the effect of TMZ on the pyruvate dehydrogenase activity, which was similar to that on lactate dehydrogenasc release. After 10 minutes of ischemia, 10 p.J4 TMZ completely blocked any activation of pyruvate dchydrogenase, whereas the activation after longer periods of ischemia was less with TMZ. Whether this effect is related to the protective mechanism of TMZ is not clear, since it will not favor glucose oxidation, but it may well allow fatty acid oxidation and thus restore a normal control of glucose oxidation by alternate oxidizable substrates during reperfusion.

CONCLUSION We have examined the effects of TMZ on the

damage incurred to mitochondrial respiration, and changes in the regulation of pyruvate dehydroge-

A SYMPOSIUM: MANAGEMENT OF MYOCARDIAL ISCHEMIA 296

nase, following global ischemia in the perfused rat heart. Any effects accorded to TMZ were only displayed at concentrations much greater (10 fi to 1 nN) than those obtained in the plasma during TMZ therapy (approximately 0.15 @I). Indeed, in a model of low-flow ischemia in the rat heart protection against cardiac contracture was ob- tained at 1 w.’ We have observed a similar protection by 1 fl TMZ in a rabbit heart low-flow ischemia model in which there is no mitochondrial damage (unpublished observation). As the develop- ment of contracture during ischemia is thought to be due to effects on the glycolytic flux,O it does appear that TMZ acts via a metabolic mechanism. Whether this is related to our observation of a better control of pyruvate dehydrogenase activa- tion by fatty acids in the presence of TMZ remains to be demonstrated.

1. Lxandy N, Martin J, Rassi A. Anti-ischemic effect of trimetazidinc 31-P.

NMR spectroscopy in the isolated rat heart. ,4rch Iru fham~~c@n 19M;Wh:YT

110.

2. Renaud JF. Internal pH, Na’, and Ca:+ regulation by trimetazidine during

cardiac cell necrosis. Canlroww Drug Thn 198X:l:h77-686.

3. Guamieri C. Muxxti C. Beneficial effects of trimetazidine on mitrochon-

drial function and superwide production in the cardiac muscle of monocmtaline-

tre&zd rats. BwcIwn l’hamracd 19&%37:-. 4. Guamieri C, Muscari C. Effecl of trimrtazidine on mitcchondrial function

and oxidattie damage during reperfusion of ischemic hypenrophied rat mywar-

dium. pharmacolqy 1993:46:324-331.

5. Veitch K, Hombnxckx A. Caucheteux D. Pouleur H, Hue L. Cl&al isch-

acmia induces a hipha\ic rcsponw of the mitochondrial respiratory chain.

Rwchem J 19YlK281:7W71.5. 6. Veitch K, Hue L. Flunarizine and cinnarizinc inhibit mitochondrial com-

plexer I and II: posihle implication for Parkinwnism. MO/ f’hurmacol IWW5:

15%163.

7. Bouchet FR, Hearse DJ. Opie LH. Effects of ttimetazidirw on ischemic

wntraaure in isolated-perfused rat hearts. I (hrr&w.sc /Ywmoc~~ 1994;24:l.

8. McCnrmack JO. Edgell NJ, Denton RM. Studies on the intwactions of

Ca:’ and pyruvate in the regulation of rat heart pymwte dehydrcgenase

activity. Bwhem J 1%2;202:41W27. 9. Owen P, Dennis S. Opie LH. Gluuw flux rate regulates omet of ischemic

contracture in &b&y underperfused rat heats. Circ Ra 199066~~354.

308 THE AMERICAN JOURNAL OF CARDIOLOGY VOLUME 76 AUGUST 24, 1995