Inhibition of brain mitochondrial respiration by dopamine: involvement of H2O2 and hydroxyl radicals but not glutathione–protein–mixed disulfides


Address correspondence and reprint requests to Gail D. Zeevalk, Department of Neurology, UMDNJ-Robert Wood Johnson Medical School, Building UBHC, Rm 401D, 675 Hoes Lane, Piscataway, NJ 08854, USA. E-mail:


Examination of the downstream mediators responsible for inhibition of mitochondrial respiration by dopamine (DA) was investigated. Consistent with findings reported by others, exposure of rat brain mitochondria to 0.5 mm DA for 15 min at 30°C inhibited pyruvate/glutamate/malate-supported state-3 respiration by 20%. Inhibition was prevented in the presence of pargyline and clorgyline demonstrating that mitochondrial inhibition arose from products formed following MAO metabolism and could include hydrogen peroxide (H2O2), hydroxyl radical, oxidized glutathione (GSSG) or glutathione–protein mixed disulfides (PrSSG). As with DA, direct incubation of intact mitochondria with H2O2 (100 µm) significantly inhibited state-3 respiration. In contrast, incubation with GSSG (1 mm) had no effect on O2 consumption. Exposure of mitochondria to 1 mm GSSG resulted in a 3.3-fold increase in PrSSG formation compared with 1.4- and 1.5-fold increases in the presence of 100 µm H2O2 or 0.5 mm DA, respectively, suggesting a dissociation between PrSSG formation and effects on respiration. The lack of inhibition of respiration by GSSG could not be accounted for by inadequate delivery of GSSG into mitochondria as increases in PrSSG levels in both membrane-bound (2-fold) and intramatrix (3.5-fold) protein compartments were observed. Furthermore, GSSG was without effect on electron transport chain activities in freeze–thawed brain mitochondria or in pig heart electron transport particles (ETP). In contrast, H2O2 showed differential effects on inhibition of respiration supported by different substrates with a sensitivity of succinate > pyruvate/malate > glutamate/malate. NADH oxidase and succinate oxidase activities in freeze–thawed mitochondria were inhibited with IC50 approximately 2–3-fold higher than in intact mitochondria. ETPs, however, were relatively insensitive to H2O2. Co-administration of desferrioxamine with H2O2 had no effect on complex I-associated inhibition in intact mitochondria, but attenuated inhibition of rotenone-sensitive NADH oxidase activity by 70% in freeze–thawed mitochondria. The results show that DA-associated inhibition of respiration is dependent on MAO and that H2O2 and its downstream hydroxyl radical rather than increased GSSG and subsequent PrSSG formation mediate the effects.

Abbreviations used:



3,4 dihydroxyphenylacetaldehyde


electron transport chain


electron transport particles


reduced glutathione


oxidized glutathione


monoamine oxidase


Parkinson's disease


glutathione–protein–mixed disulfides


reactive oxygen species.

Parkinson's disease (PD) is the most common chronic neurodegenerative movement disorder affecting older adults. PD is characterized by a loss of dopaminergic striatal projection neurons in the substantia nigra pars compacta in association with an increased turnover of dopamine (DA) within surviving neurons (Hornykiewicz and Kish 1986). Although the cause(s) of sporadic PD remains enigmatic, there is evidence to suggest that impairment of mitochondrial bioenergetics may assume an important contributory role in the neurodegenerative process. The major evidence supporting this hypothesis includes: (i) a 35% decrease in mitochondrial complex I activity in the substantia nigra pars compacta from post-mortem brains of subjects with PD (Mizuno et al. 1989; Schapira et al. 1989, 1990); and in blood platelets (Parker et al. 1989) and skeletal muscle (Blin et al. 1994) of patients with PD; (ii) induction of a parkinsonian syndrome in rodents, monkeys, and humans (Davis et al. 1979; Burns et al. 1983; Langston et al. 1983; Heikkila et al. 1984) via inhibition of complex I activity with the neurotoxic metabolite of 1-methyl-1,2,5,6-tetrahydropyridine (MPTP), 1-methyl-4-phenylpyridinium (MPP+); and (iii) induction of neuropathic changes in rodent brain that resemble those seen in PD with chronic administration of the insecticide rotenone (Betarbet et al. 2000; Helmuth 2000), a potent complex I inhibitor.

Despite these findings, the pathophysiologic mechanisms underlying the complex I defect observed in patients with PD remains elusive. Cybrid technology provides evidence that complex I defects have their origin in the mitochondrial genome as cybrids containing mitochondrial DNA from patients with PD manifest similar abnormalities in complex I function (Swerdlow et al. 1996). Alternatively, it has been postulated that the vulnerability of dopamine neurons to neurodegeneration in PD may arise as a result of intrinsic mitochondrial biochemistry related to production of reactive oxygen species (ROS) and oxidative damage related to DA metabolism (Cohen et al. 1997; Cohen and Kesler 1999). Hydrogen peroxide (H2O2) is normally formed during basal mitochondrial respiration as a result of diverting reducing equivalents, normally directed to produce water, to the production of superoxide with subsequent disproportionation to H2O2 (Chance et al. 1979). In DA neurons, however, this pathway constitutes only a portion of H2O2 production overall as a large amount derives from oxidative deamination of DA by the flavoprotein monoamine oxidase (MAO; Sinet et al. 1980; Hauptmann et al. 1996). Regardless of the underlying mechanism for abnormalities in mitochondrial respiration, turnover of intracellular DA may enhance the oxidative environment in the cell and consequently contribute cooperatively along with other factors to damage DA neurons.

The toxicity of H2O2 in part arises from its ability to induce oxidative damage to proteins involved in energy production directly as H2O2 or through its conversion into hydroxyl radicals via Fenton chemistry (Halliwell 1989; Lan and Jiang 1997). As mitochondria are lacking in catalase, mitochondrial-derived H2O2 is normally removed by its enzymatic conversion into water by glutathione peroxidase. The accompanying increase in oxidized glutathione (GSSG) can then be reconverted back to reduced glutathione (GSH) by the NADPH-dependent enzyme glutathione reductase. Under conditions of enhanced oxidative stress or increased DA turnover, however, levels of GSSG can increase many-fold (Spina and Cohen 1989; Cohen et al. 1997) promoting glutathione–protein–mixed disulfide (PrSSG) formation (Brigelius et al. 1982, 1983; Bellomo et al. 1987; Rokutan et al. 1994; Shivakumar et al. 1995; Seres et al. 1996; Cohen and Kesler 1999). In this reaction, reversibly catalyzed by thioltransferase, the sulfur atoms in the disulfide bond of GSSG are susceptible to nucleophilic substitution by protein cysteinyl–thiol residues on proteins (Ziegler 1985; Gravina and Mieyal 1993). Cohen and colleagues proposed a novel mechanism for inhibition of mitochondrial respiration resulting from increased DA metabolism via MAO, i.e. an increase in the glutathionylation of proteins to form glutathione–protein–mixed disulfides.

In support of this, Cohen and colleagues reported a dose-dependent inhibition of brain mitochondrial respiration in the presence of the MAO-B linked substrate DA and mixed MAO-A/B substrate tyramine with parallel increases in levels of GSSG and PrSSG (Cohen et al. 1997, 1999). As electron transport chain (ETC) activity is highly dependent on reduced protein cysteinyl–thiol groups (Gutman et al. 1970), it was proposed that inhibition of respiration arose from increases in PrSSG formation, resulting in inactivation of critical thiol residues in proteins of the ETC. The purpose of the present study was to expand on these findings to further characterize the involvement of PrSSG formation in the inhibition of mitochondrial respiration. The approach in this study was to examine the direct effects of GSSG versus H2O2 exposure on respiration in intact rat brain mitochondria and on the activities of different electron transport components of the ETC in freeze–thawed mitochondria and isolated inner membrane preparations from pig heart to determine the downstream mediators responsible for the observed inhibition of mitochondrial respiration in association with DA oxidation by MAO.

Materials and methods

All in vitro experiments were conducted in rats from Charles River Laboratories (Wilmington, MA, USA) in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals. All procedures were approved by the local Animal Care Committee. Rats were housed in pairs at 20–22°C on a 12-h light–dark cycle with food and water available ad libitum.

Isolation of mitochondria and ETP preparation

Mitochondria were isolated from the brains of Sprague–Dawley female rats following the procedure of (Clark and Nicklas 1970). Briefly, brains were homogenized in a glass homogenizer in 20 volumes of ice-cold isolation buffer containing: 0.225 m mannitol, 0.075 m sucrose, 5 mm MOPS and 1.0 mm EGTA, pH 7.4 The crude homogenate was centrifuged at 1800 g for 3 min, the pellet discarded and the supernatant centrifuged at 12 200 g for 8 min. The resulting pellet was resuspended in a minimal volume of 3% Ficoll, carefully layered over 6% Ficoll made up in isolation medium and centrifuged at 11 500 g for 25 min. The hard, brown pellet containing mitochondria was then resuspended in the above buffer at a volume of approximately 15 mg/mL. Electron transport particles (ETPs) were prepared from isolated pig heart mitochondria as described previously (Gluck et al. 1994). For all experiments, each separate mitochondrial preparation represents an n of one.

Measurement of O2 consumption

Oxygen consumption was carried out at 30°C in a closed chamber containing a Clark type O2 electrode hooked up to a YSI model 5300 oxygen monitor and XY chart recorder (YSI Inc., Yellow Springs, OH, USA). Mitochondria were suspended in 1 mL of incubation buffer containing: 0.095 m KCl, 0.075 m mannitol, 0.025 m sucrose, 0.005 m KH2PO4, 0.02 m Tris–HCl, 0.001 m EGTA, pH 7.4 at a concentration of 0.3 mg/mL. Pyruvate, glutatmate, malate, or succinate (potassium salts) when used in the assay were added to a final concentration of 10 mm. O2 consumption under resting conditions (no ADP, state 2) was monitored for 5 min, followed by the addition of ADP (0.25 mm final concentration) and measurement of state-3, active respiration. The mean respiratory control index (state-3/state-4) for mitochondrial preparations using pyruvate/glutamate/malate as substrate was 9.6 ± 3.7 (RCI ± SD, n = 8) and 2.0 ± 0.2, n = 10 when succinate was used as substrate. Pre-incubation with DA, GSSG or buffer alone (control) was carried out for 15 min and in the case of H2O2 or buffer alone for 5 min in the dark at 30° prior to the addition of ADP.

Measurement of glutathione–protein–mixed disulfides

Measurement was described by us previously (Ehrhart and Zeevalk 2001) and was modified from Hiranruengchok and Harris (1995). Glutathione–protein–mixed disulfides were measured as liberated glutathione and expressed as glutathione equivalents (nmol/mg protein). Protein was precipitated by addition of perchloric acid (0.4 N final concentration) and pelleted by centrifugation (16 000 g, 15 min). The protein pellet was resuspended in buffer containing 25 mm sodium pyrophosphate (pH 8.4), 1 mm dithiothreitol, 5 mm EDTA, 2.5 mm glucose-6-phosphate, 0.2 mm NADP+, 2 µg/mL glucose-6-phosphate dehydrogenase and 4 µg/mL glutathione reductase and incubated for 30 min at 37°C. After cooling, protein was again precipitated by perchloric acid and the protein pelleted by centrifugation. Liberated glutathione in the supernatant was measured by HPLC.

Oxidized and reduced glutathione measurement

Total glutathione was measured by HPLC with fluorescent detection as described previously for amino acids (Ehrhart and Zeevalk 2001). Samples were neutralized to pH 5 with K2CO3, derivitized with o-phthalaldehyde and separated on a Beckman 5 µm C18 column. Quantification was by comparison with known standards. For oxidized glutathione, at the time of collection, samples were reacted immediately with 2-vinylpyridine (2%) in 6% triethanolamine buffer as described by (Griffith 1980). Reduced glutathione was calculated by subtraction of GSSG from total glutathione.

Electron transport chain complex activities

NADH oxidase and succinate oxidase activities were measured as the rate of O2 consumption in freeze–thawed mitochondria incubated at 30°C in the presence of NADH (0.5 mm) or succinate (10 mm), respectively. In some experiments, NADH oxidase activity was also determined spectrophotometrically by following the oxidation of NADH at 340 nm over the course of several min in a dual beam spectrophotometer (Perkin Elmer, Wellesley, MA, USA). NADH cytochrome c reductase (complex I/III), succinate cytochrome c reductase (complex II/III) and cytochrome oxidase (complex IV) activities were measured using standard spectrophotometric procedures as described by Schapira et al. (1990). Rotenone (1 µm) sensitivity of NADH oxidase activity was determined in freeze–thawed mitochondrial preparations and was found to be inhibited by 95 ± 1.5% (% inhibition ± SD, n = 3) demonstrating that the majority of NADH oxidase activity was mediated via NADH dehydrogenase (complex I). Protein was measured by the procedure of Lowry et al. (1951).

Statistical analysis

For comparisons between two groups, an unpaired, two-tailed Student's t-test was used. Multiple comparisons were carried our using mnova with Tukey post-test (Instat, GraphPad Software, San Diego, CA, USA).


DA inhibits brain mitochondrial respiration by a MAO-dependent mechanism

Rat brain mitochondria were exposed to DA for 15 min during resting respiration in the presence of pyruvate/glutamate/malate (10 mmnova each). Active state-3 respiration was stimulated by the addition of ADP (0.25 mmnova) and O2 consumption was monitored for several minutes. The mean rate of O2 consumption in controls was 143 ± 28 ng atoms O2/min/mg protein. O2 consumption was inhibited by 19.5 ± 4.3% of control ± SEM with 500 µmnova DA and 24 ± 3.0% of control with 1 mmnova DA (n = 3). Incubation with DA in the presence of the MAO inhibitors pargyline plus clorgyline (2 µmnova each), significantly attenuated the inhibitory effects of DA on O2 consumption. In the presence of 500 µmnova DA, MAO inhibitors completely abolished suppression of respiration by DA (n = 3). At 1 mmnova DA plus MAO inhibitors (n = 2), respiration rates were 93% of control and were not statistically different from controls.

The effects of DA on respiration are mediated by H2O2 and not by GSSG or PrSSG

Metabolism of DA by MAO can generate a number of potential mediators that could interfere with mitochondrial proteins to inhibit respiration (Fig. 1); i.e., H2O2, hydroxyl radical, GSSG or PrSSG. To determine which events might be important in mediating the effects of DA on respiration, mitochondria were incubated with 1 mmnova GSSG (free acid, 15 min, 30°C) or 100 µmnova H2O2 (5 min, 30°C) prior to the addition of ADP and measurement of O2 consumption. Peroxide had a robust inhibitory effect on pyruvate/glutamate/malate-supported respiration (Fig. 2 and Table 1). Respiration rates with 100 µmnova H2O2 were 56 ± 3.3% of control ± SEM. In contrast, GSSG did not alter O2 consumption. In order to determine if treatment of rat brain mitochondria resulted in an increase in the glutathionylation of mitochondrial proteins, mitochondria were incubated with either 100 µmnova H2O2, 500 µmnova DA or 1 mmnova GSSG in the presence of pyruvate, glutamate and malate for 30 min at 30°C and the amount of PrSSG formation determined. H2O2 and DA increased mixed disulfides by 60 and 53%, respectively (Table 1). The mean reduction in respiration by H2O2 and DA was 44% and 22%, respectively. Incubation with 1 mmnova GSSG increased mixed disulfides to a much greater extent than H2O2 or DA, but did not cause an inhibition of respiration. The amount of mixed disulfide formation was not different in the presence or absence of ADP (data not shown).

Figure 1.

Schematic of the possible downstream mediators of MAO-catalyzed inhibition of respiration by dopamine. Metabolism of dopamine via MAO yields peroxide, ammonia and the corresponding aldehyde. Inhibition of respiration could occur by a direct action of H2O2 (A), through formation of hydroxyl radical (OH.) radical by Fenton chemistry (C) or through an increase in glutathione–protein–mixed disulfide formation due to removal of H2O2 by GSH–peroxidase plus GSH, formation of oxidized glutathione (GSSG) and subsequent formation of glutathione–protein–mixed disulfide (B).

Figure 2.

Representative trace of O2 consumption in isolated rat brain mitochondria with pyruvate/glutamate/malate as substrates following incubation with 100 µmnova H2O2 for 5 min or 1 mmnova GSSG for 15 min at 30°C. Active state-3 respiration was stimulated by the addition of 0.25 mmnova ADP (arrow).

Table 1.  Inhibition of mitochondrial respiration does not correlate with an increase in mitochondrial glutathione–protein–mixed disulfides
ConditionGSH–protein–mixed disulfide
n(-fold change ± SEM)n(% control ± SEM)
  1. Rat brain mitochondria were isolated as described in Methods and incubated with either dopamine (DA), H2O2 or oxidized glutathione (GSSG) at the above indicated concentrations for 30 min at 30°C. At the end of incubation, mitochondria were analyzed for glutathione–protein–mixed disulfide formation. In separate incubations, from the same preparations, mitochondria were exposed to the above compounds using incubation conditions as described in the legends to Figs 2 and 3 and the effects on pyruvate/glutamate/malate supported respiration was determined. aDifferent from control, p < 0.05.

Control121.00 ± 0.0712100 ± 6.1
100 µmnova H2O2 61.60 ± 0.15a 6 56 ± 3.3a
500 µmnova DA 31.53 ± 0.19a 3 78 ± 2.6a
1 mmnova GSSG 33.30 ± 0.57a 3108 ± 4.4

To insure that extramitochondrial GSSG was taken up by mitochondria and had access to intramitochondrial proteins, mitochondria were incubated with 1 mmnova GSSG (30 min, 30°C), washed, lysed and intramatrix GSH and GSSG as well as PrSSG formation in soluble matrix proteins and membrane bound proteins was determined. As shown in Table 2, intramitochondrial GSSG levels increased 26-fold, whereas GSH levels increased 4-fold. HPLC analysis of freshly made GSSG showed no measurable reduced glutathione contamination. The glutathionylation of insoluble membrane-bound and soluble matrix proteins increased by two- and 3.5-fold, respectively, demonstrating access of GSSG to intramitochondrial proteins. To further provide adequate access of GSSG to mitochondrial proteins, and to more directly examine the effects of GSSG on ETC complex activities, freeze–thawed mitochondria were exposed to 1 mmnova GSSG for 15 min, 30°C and the effects on NADH oxidase activity monitored by O2 consumption or ETC complex activities monitored with spectrophotometric assays, were determined. NADH oxidase activity, which measures the combined activities of complexes I, III, and IV, in freeze–thawed mitochondria was not inhibited by GSSG exposure (97 ± 2.4% of control ± SEM, n = 3). Likewise, enzymatic determination of NADH cytochrome c reductase, succinate cytochrome c reductase or cytochrome oxidase was unaffected by GSSG (Fig. 3). Similar results were observed with inverted inner mitochondrial membranes (ETPs) from pig heart (data not shown).

Table 2.  Effect of extramitochondrial GSSG on intramitochondrial GSH, GSSG and glutathione–protein–mixed disulfides (nmol/mg protein ± SEM)
  1. Isolated rat brain mitochondria were incubated in the presence or absence of 1 mmnova GSSG for 30 min at 30°C. At the end of incubation, the mitochondria were washed, lysed and the intramatrix GSH and GSSG content were measured by HPLC. Glutathione–protein–mixed disulfide formation was measured separately in soluble matrix proteins and in the total mitochondrial membrane fraction. n is from four to six determinations per condition. aDifferent from control, p < 0.05.

Control 3.01 ± 0.500.13 ± 0.060.0430.68 ± 0.080.16 ± 0.03
1 mmnova GSSG13.40 ± 2.10a3.41 ± 0.90a0.2541.43 ± 0.24a0.57 ± 0.19a
Figure 3.

Freeze–thawed mitochondria were assayed spectrophotometrically for NADH cytochrome c reductase (combined complex I/III), succinate cytochrome c reductase (combined complex II/III) and cytochrome oxidase (complex IV) activities (▮) or assayed for O2 consumption (NADH oxidase activity, □) following a 15-min incubation with 1 mmnova GSSG at 30°C. n is from three determinations per condition.

Differential effects of H2O2 on respiration and modulation by iron chelation

Brain mitochondria showed a dose-dependent decrease in respiration when incubated with varying concentrations of H2O2 for 5 min prior to measurement of O2 consumption. The potency of H2O2 to inhibit respiration in intact mitochondria was substrate specific (Fig. 4a). Succinate supported respiration was most sensitive to H2O2 exposure. For substrates that feed into complex I, respiration supported by pyruvate/malate was more sensitive to inhibition by H2O2 than that supported by glutamate/malate, whereas pyruvate/glutamate/malate was intermediate in sensitivity. The IC50 for inhibition of O2 consumption by H2O2 versus NADH oxidase and succinate oxidase were increased 2–3-fold when mitochondria were lysed by freeze–thawing (Fig. 4b). NADH or succinate oxidase activities as measured by O2 consumption in ETPs, which contain purified inner mitochondrial membranes in reverse orientation, were inhibited by H2O2 to a much lesser degree: 40% inhibition with 5 mmnova H2O2 (Fig. 4b).

Figure 4.

(a) Rat brain mitochondria were treated with different concentrations of H2O2 (10–1000 µmnova) for 5 min, 30°C and the rate of O2 consumption using different substrates to support respiration was determined. IC50 were calculated from nonlinear regression plots (GraphPad, Inplot). n is from three to five dose–response determinations per substrate. (b) Intact mitochondria, freeze–thawed mitochondria or electron transport particles (ETPs) were exposed to different concentrations of H2O2 for 5 min and the effects on O2 consumption for pyruvate/malate- or succinate-supported respiration (intact), or NADH oxidase or succinate oxidase activities (freeze–thawed or ETPs) were determined as described in Methods. n is from three to four determinations per dose–response for intact and freeze–thawed mitochondria and two for ETPs.

To determine if hydroxyl radicals were possible mediators of the effects of H2O2 on respiration and if extramitochondrial versus intramitochondrial hydroxyl radical formation could produce inhibitory effects on respiration, both intact and freeze–thawed mitochondria were incubated with H2O2 for 5 min in the presence or absence of the iron chelator desferrioxamine mesylate (1 mmnova). The effects on O2 consumption with pyruvate/glutamate/malate-supported respiration in intact mitochondria versus NADH oxidase activity in freeze–thawed preparations was determined as presented in Figs 5(a and b). Iron chelation and prevention of extramitochondrial hydroxyl radical generation via Fenton chemistry in intact mitochondria did not attenuate the inhibition of respiration by H2O2 (Fig. 5a). However, iron chelation with 1 mmnova desferrioxamine in freeze–thawed mitochondria significantly reduced the inhibitory effect of H2O2 on NADH oxidase activity (Fig. 5b). Desferrioxamine alone did not alter O2 consumption in intact mitochondria or NADH oxidase activity in freeze–thawed mitochondria.

Figure 5.

(a) Brain mitochondria were exposed to different concentrations of H2O2 in the presence or absence of 1 mmnova desferrioxamine for 5 min at 30°C and the effects on pyruvate/glutamate/malate supported respiration was determined. (b) Rat brain mitochondria were lysed by freeze–thawing and exposed to sufficient H2O2 for 5 min at room temperature to produce approximately 50% inhibition of NADH oxidase activity (400–800 µmnova). Parallel incubations were carried out in the presence of 1 mmnova desferrioxamine. NADH oxidase activity was measured spectrophotometrically by following the oxidation of NADH to NAD in a dual beam spectrophotometer. n is from three determinations. *Different from H2O2 alone.


A characteristic of PD is a compensatory increase in the turnover of DA within surviving neurons (Hornykiewicz and Kish 1986). DA turnover is largely achieved by the flavoprotein MAO-B located on the outer mitochondrial membrane (Ragan et al. 1987). MAO oxidatively deaminates DA to 3,4-dihydroxyphenylacetaldehyde (DOPAL), ammonia and H2O2. Mitochondria remove peroxide via glutathione peroxidase using GSH as reductant thus leading to increases in GSSG levels. The importance of DA turnover via MAO in increasing GSSG levels in DA neurons has been demonstrated (Berman and Hastings 1999; Cohen and Kesler 1999). GSSG is susceptible to nucleophilic attack by cysteinyl thiol residues in proteins resulting in PrSSG formation. Studies have shown that PrSSG can lead to modulation of enzyme function (Baba et al. 1978; Gilbert 1982; Ziegler 1985; Quio et al. 2000). ETC complex activities are readily inhibited by thiol-modifying agents (Gutman et al. 1970) and could thus be subject to glutathionylation. Increased DA turnover by MAO and increased GSSG (Cohen and Kesler 1999) occur in parallel with an increase in mitochondrial PrSSG formation. Based on these findings, it has been hypothesized (Cohen and Kesler 1999) that GSSG and increased glutathionylation of mitochondrial proteins contributed to the inhibition of the ETC by MAO-dependent oxidation of DA.

The present study demonstrates that GSSG and increased glutathionylation of mitochondrial proteins does not alter the activities of the ETC complexes in either intact or freeze–thawed mitochondria or in ETPs and therefore does not support this pathway for DA-mediated inhibition of mitochondrial respiration. This conclusion is based on the following evidence. Direct exposure of intact mitochondria to GSSG had no effect on respiration. It is important to note that the free acid of GSSG was used for these studies, since the sodium salt of oxidized glutathione will inhibit O2 consumption due to the well known effects of high sodium on mitochondrial respiration. Oxidative stress and increased GSSG has been shown to increase PrSSG formation in whole cells (Brigelius et al. 1982, 1983; Bellomo et al. 1987) and in intact mitochondria (Benard and Balasubramanian 1995; Olafsdottir and Reed 1988; Ravindranath and Reed 1990). Consistent with this, exposure of mitochondria to DA, H2O2 or GSSG resulted in a significant increase in PrSSG, but in contrast with DA or H2O2, GSSG did not inhibit respiration suggesting a dissociation between effects on respiration and mitochondrial PrSSG formation. It is possible that the glutathionylation of mitochondrial proteins observed in the intact mitochondria were the result of PrSSG formation of proteins in the outer mitochondrial membranes and that GSSG did not access the intramitochondrial space. Two approaches were used to test this possibility. In one study, mitochondria were incubated with GSSG and intramitochondrial concentrations of GSH and GSSG as well as PrSSG formation in membrane-bound and soluble matrix proteins were determined. In a second approach, the effects of GSSG on ETC activity in lysed, freeze–thawed mitochondria were determined. Intact mitochondria incubated with GSSG increased intramitochondrial GSSG and GSH as well as the glutathionylation of membrane-bound and soluble matrix proteins indicating that extramitochondrial GSSG is able to be taken up by mitochondria. Thus, some of the GSSG is converted back to GSH via mitochondrial glutathione reductase. The increase in GSSG also leads to covalent modification of cysteinyl–thiols of proteins to form a mixed disulfide. Additionally, NADH oxidase activity, that evaluates ETC activities I, III and IV in lysed, freeze–thawed mitochondria was also unaltered by GSSG exposure. To further evaluate the effects of GSSG and PrSSG on ETC activities, enzymatic spectrophotometric assays were used to monitor ETC activities in freeze–thawed mitochondria. Under these conditions, GSSG, was found to be without effect. The above data strongly argue that the increase in GSSG caused by DA turnover via MAO does not contribute to the inhibitory effects of DA on mitochondrial respiration.

Although GSSG was without effect on mitochondrial respiration, H2O2 was a fairly potent inhibitor of respiration in intact mitochondria. Five-minute exposures of mitochondria to H2O2 led to dose-dependent inhibition of respiration supported by complex I and II substrates. The potency of H2O2 in inhibiting respiration in intact mitochondria was dependent on the substrates used. Succinate supported respiration was the most sensitive, pyruvate/malate intermediate in sensitivity and glutamate/malate, the least sensitive with regard to the inhibitory effects of H2O2. Whilest it might be argued that SDH/complex II is more sensitive to the effects of H2O2 versus complex I, the differences in IC50 for H2O2 versus pyruvate/malate and glutamate/malate, both substrates that feed into complex I, suggest that the effects of H2O2 are not limited to ETC complexes, but also involve other components of the mitochondrial machinery. The greater sensitivity of pyruvate/malate versus glutamate/malate to H2O2 is similar to findings reported by Sims et al. (2000). The pyruvate dehydrogenase complex is subject to oxidative inactivation (Tabatabaie et al. 1996) and could be a possible reason for the sensitivity of pyruvate/malate supported respiration to H2O2. However, H2O2 was reported not to inhibit pyruvate dehydrogenase activity in intact mitochondria (Sims et al. 2000). Thus the reasons underlying the differences in sensitivity of pyruvate/malate versus glutamate/malate to H2O2 remain to be elucidated. In contrast with our finding of a greater susceptibility of succinate supported respiration to H2O2, Sims et al. (2000) did not observe an inhibition of respiration supported by succinate. While the reasons for these differences are not clear, it should be noted that DA was found to inhibit O2 consumption supported by succinate and was dependent on MAO metabolism, suggesting mediation by H2O2 (Berman and Hastings 1999; Cohen and Kesler 1999).

NADH and succinate oxidase activities in freeze–thawed mitochondria were also inhibited by H2O2 with 2–3-fold higher concentrations needed to obtain a similar degree of inhibition compared with intact mitochondria. Because NADH and succinate oxidase activities monitor the combined activities of the ETC complexes, independent of transporters, Krebs cycle enzymes, ATPase, etc., the greater IC50 values are consistent with the argument that the inhibitory actions of H2O2 are mediated by direct effects on the ETC components as well as other mitochondrial constituents. The inhibition of succinate oxidase activity in freeze–thawed mitochondria is also in agreement with an inhibitory effect of H2O2 on succinate supported O2 consumption in intact mitochondria.

The findings of MAO-dependent inhibition of brain mitochondrial respiration reported here are in agreement with those of others (Berman and Hastings 1999; Cohen and Kesler 1999) in that a short pre-incubation with DA led to an inhibition of state-3 respiration. In the present study, 500 µmnova DA for 15 min inhibited pyruvate/glutamate/malate supported respiration by 20% consistent with the degree of inhibition of pyruvate/malate supported respiration by a 15-min incubation with 500 µmnova tyramine (24%) reported by Cohen and Kesler (1999). Glutamate/malate-supported respiration in intact rat brain mitochondria was reported to be inhibited by 24% with a 5-min exposure to 100 µmnova DA (Berman and Hastings 1999). The modest increase in sensitivity in this study is likely due to the mitochondrial incubations being carried out in a balanced salt solution lacking either mannitol (current study) or BSA (Cohen and Kesler 1999), both used to maintain the integrity and coupling of mitochondria. Despite the different assay conditions, the effects and potency of DA on respiration supported by substrates that feed into complex I in intact mitochondrial preparations were remarkably consistent. Also consistent were the findings that inhibition of active respiration by DA could be blocked by MAO inhibitors (Berman and Hastings 1999; Cohen and Kesler 1999). In contrast, reports of inhibition of respiration by DA using tissue homogenates or disrupted mitochondria have produced equivocal results (Przedborski et al. 1993; Morikawa and Nakagawa-Hattori Ya 1996; Ben-Shachar et al. 1997).

Unlike the fairly potent inhibition of respiration or ETC activity in intact or freeze–thawed mitochondria observed in the present study, isolated ETPs were relatively insensitive to the effects of H2O2 requiring mmnova concentrations to achieve modest inhibitory effects. This finding is in agreement with Zhang et al. (1990). In the later study, beef heart ETP activities for NADH dehydrogenase, NADH oxidase, succinate dehydrogenase, and succinate oxidase were rapidly inactivated by hydroxyl radicals, but were relatively insensitive to direct H2O2 exposure. This would suggest that the large discrepancy between the response of ETPs and mitochondria with regard to H2O2 may be due to formation of hydroxy radicals from H2O2 rather than H2O2 itself. Purified ETP preparations are likely to have a reduced free iron content as compared with intact or freeze–thawed mitochondrial preparations, thus reducing Fenton chemistry. This hypothesis was examined by exposing either intact or freeze–thawed mitochondria to H2O2 in the presence of desferrioxamine, a chelator of iron. Consistent with findings of others (Sims et al. 2000), desferrioxamine was not able to prevent the inhibitory effects of H2O2 in intact mitochondria. Because desferrioxamine does not readily cross the inner mitochondrial membrane in intact mitochondria (Sims et al. 2000), this indicates that extra-mitochondrial hydroxyl radical formation from H2O2 is not disruptive to mitochondrial respiration. In contrast, iron chelation in freeze–thawed mitochondria significantly attenuated the inhibition of NADH oxidase by H2O2 pointing towards intramitochondrial hydroxyl radical formation from H2O2 as responsible for respiratory inhibition. These findings further suggest that DA metabolism via MAO may lead to increased levels of intramitochondrial hydroxyl radicals that in turn could perturb mitochondrial function.

In summary, these findings confirm reports by others (Berman and Hastings 1999; Cohen and Kesler 1999) that DA turnover via MAO can lead to inhibition of mitochondrial respiration. Elucidation of the downstream mediators of this effect strongly argue against increased GSSG and PrSSG formation, and instead, provide evidence that intramitochondrial hydroxyl radicals formed by Fenton chemistry from H2O2 generated during MAO metabolism serves as a major contributor. Of interest, is a recent report (Kristal et al. 2001), to show that the aldehyde intermediate DOPAL could promote mitochondrial permeability transition in de-energized mitochondria. This raises the intriguing possibility that DA metabolism via MAO could lead to a double hit on mitochondria; hydroxyl radical damage to energy transfer components leading to de-energized mitochondria and production of an aldehyde intermediate that could in turn promote permeability transition.


This work is dedicated to the memory of Dr Gerald Cohen. Dr Cohen's work has generated important insights and has done much to stimulate research in the field of Parkinson's research. He will be missed by his colleagues. This work was supported by a Career Development Award from the Department of Veterans Affairs and a Presidential Early Career Award for Scientists and Engineers to (MRG) and Grants from the Michael J. Fox Foundation for Parkinson's Research and Public Health Service grant NS36157 to (GDZ).