Dopamine Oxidation Alters Mitochondrial Respiration and Induces Permeability Transition in Brain Mitochondria

Implications for Parkinson’s Disease


  • Preliminary reports of these findings were presented at the 27th Annual Meeting of the Society for Neuroscience, New Orleans, LA, U.S.A., October 25-30, 1997.

  • Abbreviations used : CsA, cyclosporin A ; DA, dopamine ; H2O2, hydrogen peroxide ; MAO, monoamine oxidase ; O2.-, superoxide anion ; 6-OHDA, 6-hydroxydopamine ; PD, Parkinson’s disease ; PTP, permeability transition pore ; ROS, reactive oxygen species ; SOD, superoxide dismutase ; TMPD, N,N,N’,N’-tetramethyl-p-phenylenediamine.

Address correspondence and reprint requests to Dr. T. G. Hastings at Department of Neurology, S-505 Biomedical Science Tower, University of Pittsburgh, Pittsburgh, PA 15261, U.S.A.


Abstract : Both reactive dopamine metabolites and mitochondrial dysfunction have been implicated in the neurodegeneration of Parkinson’s disease. Dopamine metabolites, dopamine quinone and reactive oxygen species, can directly alter protein function by oxidative modifications, and several mitochondrial proteins may be targets of this oxidative damage. In this study, we examined, using isolated brain mitochondria, whether dopamine oxidation products alter mitochondrial function. We found that exposure to dopamine quinone caused a large increase in mitochondrial resting state 4 respiration. This effect was prevented by GSH but not superoxide dismutase and catalase. In contrast, exposure to dopamine and monoamine oxidase-generated hydrogen peroxide resulted in a decrease in active state 3 respiration. This inhibition was prevented by both pargyline and catalase. We also examined the effects of dopamine oxidation products on the opening of the mitochondrial permeability transition pore, which has been implicated in neuronal cell death. Dopamine oxidation to dopamine quinone caused a significant increase in swelling of brain and liver mitochondria. This was inhibited by both the pore inhibitor cyclosporin A and GSH, suggesting that swelling was due to pore opening and related to dopamine quinone formation. In contrast, dopamine and endogenous monoamine oxidase had no effect on mitochondrial swelling. These findings suggest that mitochondrial dysfunction induced by products of dopamine oxidation may be involved in neurodegenerative conditions such as Parkinson’s disease and methamphetamine-induced neurotoxicity.

In Parkinson’s disease (PD), the cause of the degeneration of dopaminergic neurons of the substantia nigra is unknown, but evidence suggests that oxidative stress is involved (for review, see Fahn and Cohen, 1992). One source of oxidative stress that is unique to dopaminergic neurons is the presence of dopamine (DA) itself, as DA can form reactive oxygen species (ROS) and quinones through two different pathways. First, DA is metabolized via monoamine oxidase (MAO) to produce hydrogen peroxide (H2O2) and dihydroxyphenylacetic acid (Maker et al., 1981). H2O2, if not reduced by cellular antioxidant mechanisms such as GSH and GSH peroxidase, can react with transition metals such as iron to form hydroxyl radical (Halliwell, 1992). This molecule will immediately react with lipids, DNA, and susceptible amino acids in proteins, thus causing cellular damage (Halliwell, 1992). Second, the catechol ring of DA can undergo oxidation to form DA quinone and ROS such as H2O2 and superoxide anion (O2•-) in a reaction that can occur either spontaneously in the presence of transition metals or enzymatically (Graham, 1978 ; Hastings, 1995). The DA quinone is electron-deficient and reacts readily with cellular nucleophiles such as sulfhydryl groups on free cysteine, GSH, and cysteinyl residues in proteins (Tse et al., 1976 ; Graham, 1978). The reaction between the DA quinone and sulfhydryl groups leads to covalent modification of protein and free thiols, forming cysteinyl-DA conjugates (Tse et al., 1976 ; Graham, 1978 ; Fornstedt et al., 1990 ; Hastings and Zigmond, 1994). Because free thiols are important antioxidants in cells and protein cysteinyl residues often play critical roles in protein function, alterations of either free or protein thiols could lead to cellular toxicity.

DA is known to be toxic both in vitro (Graham, 1978 ; Michel and Hefti, 1990) and in vivo (Filloux and Townsend, 1993 ; Hastings et al., 1996), and we have shown that the formation of cysteinyl-DA conjugates correlates with DA-induced neurotoxicity (Hastings et al., 1996). In addition, we have shown that DA oxidation products can inhibit the function of specific proteins, the DA and glutamate transporters (Berman et al., 1996 ; Berman and Hastings, 1997), and others have recently reported similar effects on the activities of tryptophan hydroxylase (Kuhn and Arthur, 1998) and tyrosine hydroxylase (Xu et al., 1998).

The protein targets that are critical to the toxicity induced by DA are not yet known, but likely candidates include many of the proteins important in mitochondrial processes. The critical role of mitochondria for cellular survival is well known, and mitochondrial dysfunction has recently been elucidated as an essential target in the induction of apoptosis as well as in excitotoxic neuronal death (Deckwerth and Johnson, 1993 ; Vayssiére et al., 1994 ; Zamzami et al., 1995a, b ; Petit et al., 1995 ; Liu et al., 1996 ; Schinder et al., 1996 ; Susin et al., 1996 ; White and Reynolds, 1996 ; Ellerby et al., 1997). These findings have led to a focus on potential contributions of mitochondrial dysfunction to neurodegenerative diseases (see Bowling and Beal, 1995). Mitochondria are of particular interest in PD, where evidence has suggested that an underlying deficit of complex I enzyme activity in the mitochondrial electron transport chain exists (Parker et al., 1989 ; Schapira et al., 1990 a, b ; Sheehan et al., 1997). Whether this plays a causative role in PD has not yet been elucidated, but it suggests that deficiencies in mitochondrial function could be involved in the degeneration of DA neurons.

Several mitochondrial processes can be disrupted by oxidants such as ROS and quinones. One such process is mitochondrial respiration, which is responsible for generating ATP through oxidative phosphorylation. Several enzymes in the electron transport chain have been shown to be inhibited following exposure to ROS or sulfhydryl-modifying agents (Kenney, 1975 ; Yagi and Hatefi, 1987 ; Zhang et al., 1990 ; Benard and Balasubramanian, 1995). Because both ROS and quinones, formed as a result of DA oxidation, are capable of modifying critical sulfhydryl groups on proteins, these electron transport enzymes may be susceptible to damage by DA oxidation products.

Another potential target of DA oxidation products is the mitochondrial permeability transition pore (PTP). The PTP is a calcium-dependent, proteinaceous pore that allows the normally impermeable inner membrane of mitochondria to become permeable to solutes of <1,500 Da. The change in membrane permeability leads to depolarization of the transmembrane potential, release of small solutes and then proteins, osmotic swelling, and a loss of oxidative phosphorylation (for review, see Gunter and Pfeiffer, 1990). Opening of the PTP has been implicated in several forms of neuronal death including apoptosis, excitotoxicity, ischemia, and toxicity due to the parkinsonian neurotoxin MPTP (Uchino et al., 1995 ; Nieminen et al., 1996 ; Packer et al., 1996 ; Schnider et al., 1996 ; White and Reynolds, 1996 ; Zamzami et al., 1996 ; Ouyang et al., 1997 ; Cassarino et al., 1998). Many oxidants and toxic quinones are known inducers of PTP opening (e.g., see Gunter and Pfeiffer, 1990). Likewise, sulfhydryl modification has been shown to induce PTP opening, and critical cysteinyl residues have been implicated in regulation of the PTP (Bernardi et al., 1992 ; Valle et al., 1993 ; Petronilli et al., 1994). Therefore, the PTP is also a potential target of ROS and quinones formed through both DA oxidation pathways.

In this study, we examined the effects of DA oxidation products on both mitochondrial respiration and the PTP, using isolated respiring brain mitochondria. We report that MAO-catalyzed oxidation of DA and production of H2O2 inhibit active mitochondrial respiration, whereas DA quinone production leads to a large increase in resting respiration, indicative of an increase in inner membrane permeability. In addition, we found that the oxidation of DA to DA quinone results in a cyclosporin A (CsA)-inhibitable increase in mitochondrial swelling, suggestive of the opening of the PTP. These effects on mitochondrial function could contribute to DA-induced toxicity and to the neurodegenerative process in PD.


Mithcondrial isolation

Brain mitochondria were isolated from male Sprague-Dawley rats (300-350 g) by the method of Rosenthal et al. (1987). This method uses 0.02% digitonin to free mitochondria from the synaptosomal fraction. In brief, one rat was decapitated, and the whole brain minus the cerebellum was rapidly removed, washed, minced, and homogenized in a Dounce glass tissue homogenizer (via six strokes each with a loose-fitting pestle and then a tight-fitting pestle) at 4°C in 10 ml of isolation medium (225 mM mannitol, 75 mM sucrose, 5 mM HEPES, 1 mM EGTA, 1 mg/ml bovine serum albumin, pH 7.4) containing 5 mg of the bacterial protease Nagarse. Single brain homogenates were brought to 30 ml, divided equally into three tubes, and then centrifuged at 2,000 g for 3 min. Pellets were resuspended to 10 ml and recentrifuged as above, and the supernatants were pooled and centrifuged in four tubes at 12,000 g for 8 min. The pellets, including the fluffy synaptosomal layer, were resuspended in two tubes to 10 ml each in isolation medium containing 0.02% digitonin and centrifuged at 12,000 g for 10 min. The brown mitochondrial pellets without the synaptosomal layer were then resuspended again in 10 ml of medium and recentrifuged at 12,000 g for 10 min. The mitochondrial pellets were resuspended in 50 μl of medium/tube and combined. Mitochondrial protein yields, determined by the method of Bradford (1976), were ~8-12 mg per rat brain. When utilized, liver mitochondria were isolated from 1.5-1.75 g of liver tissue using the identical procedure, which produced 20-25 mg of mitochondrial protein.

Mitochondrial respiration

Respiration measurements were determined polarographically with a thermostatically controlled (37°C) Clark oxygen electrode (Yellow Springs Instrument Co., Yellow Springs, OH, U.S.A.) according to the method of Rosenthal et al. (1987) in standard respiration medium containing 125 mM KCl, 2 mM K2HPO4, 1 mM MgCl2, 5 mM K-HEPES (pH 7.0), 1 mM EGTA, 5 mM glutamate, and 5 mM malate. Mitochondria (0.5 mg of protein/ml) were added to 1.6 ml of medium in a water-jacketed chamber (Gilson, Middleton, WI, U.S.A.). Active state 3 respiration rates were determined by the addition of ADP (0.25 mM), and resting state 4 respiration rates were determined after consumption of ADP and the addition of oligomycin (2 μg/ml) to inhibit ATP synthase. Rates are expressed as nanograms of oxygen atoms consumed per minute per milligram of protein and were calculated based on the solubility of oxygen in the air-saturated, temperature-equilibrated medium of 390 ng of O/ml at 37°C and 760 mm Hg. Evaluation of state 3 and state 4 rates occurred over ~3-5 min for each sample. Prior to the initiation of every experiment, respiration rates of the mitochondrial preparation were determined, and mitochondria were used for these studies when the ratio of state 3 respiration to state 4 respiration was determined to be at least 7.0, signifying healthy, well-coupled mitochondria.

For experiments examining the effects of DA oxidation products on mitochondrial respiration, mitochondria (0.5 mg of protein/ml) were incubated in medium alone or medium containing the indicated compounds for 5 min in the electrode chamber at 37°C, with air bubbled into the chamber to maintain O2 saturation. All control incubations were performed in an identical manner. At the end of the incubation period, state 3 and state 4 respiration was measured as described above. In experiments examining succinate-linked respiration, the medium contained 125 mM KCl, 2 mM K2HPO4, 1 mM MgCl2, 5 mM K-HEPES (pH 7.0), 1 mM EGTA, 5 mM succinate, and 2 μM rotenone. For experiments using ascorbate and N,N,N’,N’-tetramethyl-p-phenylenediamine (TMPD) to examine complex IV activity, mitochondria were incubated for 5 min in the standard respiration medium containing malate (5 mM) and glutamate (5 mM). After the incubation peroid, rotenone (2 μM) was added, followed by the addition of antimycin A (1 μM), ascorbate (2 mM), and TMPD (0.1 mM), and state 3 respiration was measured with the addition of ADP (0.25 mM).

Mitochondrial swelling

Mithcondrial swelling was measured spectrophotometrically (Beckman DU-640, Fullerton, CA, U.S.A.) by monitoring the decrease in absorbance at 540 nm over 10 min similar to previously described methods (Broekemeier et al., 1989). Mitochondria (1 mg of protein) were incubated in 2 ml of medium containing 213 mM mannitol, 70 mM sucrose, 3 mM HEPES (pH. 7.4), 10 mM succinate, and 1 μM rotenone. CaCl2 (70 μM) was added after 30 s, and other indicated compounds were added at 2 min. When CsA or GSH was used, it was added to the buffer prior to the addition of the mitochondria. When tyrosinase was used to oxidize DA and when 6-hydroxydopamine (6-OHDA) was used, interfering absorbance due to colored oxidative products was subtracted from measurements using blanks containing only buffer with DA and tyrosinase or buffer with 6-OHDA. Data were quantified and compared by calculating the total decrease in absorbance from 2 to 10 min.

Statistical analysis

Analyses were performed by one-way ANOVA followed by Tukey’s post hoc comparisons. A probability of p < 0.05 was considered significant. The n values reported refer to data obtained from n separate experiments.


Mitochondrial respiration

A typical measure of oxygen consumption in our isolated brain mitochondrial preparation is shown in Fig. 1. Mitochondrial respiration is conventionally classified into several states, which can be measured via oxygen consumption (Chance and Williams, 1956). State 3, termed active respiration, is defined as respiration in the presence of an oxidizable substrate and ADP and thus is a measure of the respiration that is coupled to ATP synthesis. State 4, or resting, respiration is the rate of respiration in the presence of substrate, but without ADP, and thus is a measure of the rate of respiration that is not coupled to ATP synthesis. Mean rates of active ADP-linked state 3 respiration and resting state 4 respiration were 216 ± 17 and 23 ± 2 ng of O/min/mg of protein, respectively, in untreated brain mitochondria (n = 28 ; mean ± SEM). The ratio of state 3 to state 4 can be used to evaluate the functional health of the preparation by giving an indication of the degree to which respiration is coupled to ATP synthesis. The average ratio of state 3 to state 4 was 10.2. For experiments in which liver mitochondria were used, mean state 3 and state 4 respiration was 219 ± 44 and 21 ± 5.6 ng of O/min/mg of protein, respectively, and did not differ significantly from that in isolated brain mitochondria (n = 6 ; mean ± SEM).

Figure 1.

Representative measure of baseline mitochondrial respiration. For each isolated mitochondrial preparation, oxygen consumption was first measured in isolated brain mitochondria (0.5 mg of protein/ml) with glutamate and malate as substrates, prior to any experimentation, as described in Materials and Methods. State 3 respiration was measured after addition of ADP (0.25 mM), and state 4 respiration was measured after addition of oligomycin (2 μg/ml). The rate of uncoupled respiration was recorded after the addition of FCCP (carbonyl cyanide p-trifluoromethoxyphenylhydrazone ; 150 nM). Mean rates of oxygen consumption were 216 ± 17 and 23 ± 2 ng of O/min/mg of protein for state 3 and state 4, respectively.

FIG. 1.

Effects of DA oxidation on mitochondrial respiration

We first examined the effect of DA alone on mitochondrial respiration. When mitochondria were incubated for 5 min in the control respiration buffer, mean state 3 respiration was 86 ± 5 ng of O/min/mg of protein, and mean state 4 respiration was 18 ± 3 ng of O/min/mg of protein (Fig. 2). The reduction in state 3 respiration is typical for isolated brain mitochondria after incubation periods at 37°C. When brain mitochondria were incubated for 5 min in respiration buffer containing DA (100 μM), state 3 respiration was reduced by 24% as compared with respiration after incubation in buffer alone (Fig. 2). However, state 4 respiration was unaffected by the presence of DA (Fig. 2). To determine whether the state 3 inhibition caused by DA was due to H2O2 formed during MAO-catalyzed oxidation of DA, we examined the effects of adding the nonselective MAO inhibitor pargyline (10 μM) or the H2O2-scavenging enzyme catalase (1 U/ml) to the incubation medium. We found that the addition of either pargyline or catalase during the incubation with DA completely prevented the inhibition of state 3 respiration (Fig. 2). Neither pargyline nor catalase alone had any effect on mitochondrial respiration (data not shown).

Figure 2.

Effect of DA on mitochondrial respiration. Isolated brain mitochondria were incubated for 5 min as described in MATERIALS AND METHODS in buffer alone or buffer containing DA (100 μM), DA + pargyline (Parg ; 10 μM), or DA + catalase (Cat ; 1 U/ml), and then oxygen consumption was measured, utilizing glutamate and malate as substrates. Under control conditions, mean state 3 respiration was 86 ± 5 ng of O/min/mg of protein, and mean state 4 respiration was 18 ± 3 ng of O/min/mg of protein. *Significantly less than respiration under control conditions ; **significantly greater than respiration after exposure to DA alone (mean ± SEM ; n = 5-16).

FIG. 2.

To begin to determine which enzyme complexes of the electron transport chain were affected by DA, we utilized various substrates that donate electrons at different points in the electron transport chain, thereby hoping to bypass the DA-induced inhibition. We observed that when succinate and rotenone were utilized to bypass complex I and donate electrons via complex II to complex III, the decrease in state 3 respiration following incubation with DA was still observed (Fig. 3). Under these conditions, state 3 respiration was inhibited by 37% in the presence of DA as compared with control, an inhibition that was completely prevented by pargyline. However, when ascorbate and TMPD were utilized to donate electrons directly to cytochrome c, bypassing complexes I and III, incubation with DA no longer had any effect on state 3 respiration as compared with incubation under the same conditions without DA (Fig. 3).

Figure 3.

Effect of bypassing steps in the electron transport chain on DA-induced inhibition of state 3 respiration. Mitochondria were incubated in buffer alone or buffer containing DA (100 μM) or DA + pargyline (10 μM) as described in MATERIALS AND METHODS. Succinate (5 mM) + rotenone (2 μM) were utilized to donate electrons to complex II, and ascorbate (2 mM) + TMPD (100 μM) + antimycin A (1 μM) were utilized to donate electrons to cytochrome c. *Significantly less than respiration under control conditions (mean ± SEM ; n = 3-8).

FIG. 3.

To examine whether the products resulting from oxidation of the catechol ring would alter mitochondrial respiration, we utilized the enzyme tyrosinase to directly oxidize DA to DA quinone. Mitochondria were incubated for 5 min in standard respiration medium or medium containing DA (20 or 100 μM) and a concentration of tyrosinase (200 mU/ml) that was determined to oxidize all of the DA within 2 min (Berman and Hastings, 1997). Tyrosinase alone had no effect on respiration (data not shown). Following the incubation, we found that incubation with either concentration of DA quinone had no significant effect on state 3 respiration (Fig. 4A). However, we observed a dramatic increase in state 4 respiration with both the lower and the higher DA concentrations (+202 and +280%, respectively ; Fig. 4B). Coincubation with GSH (500 μM), which can either reduce the DA quinone to DA or act as a nucleophile to bind DA, prevented the DA oxidation-induced increase in state 4 respiration at both concentrations of DA (Fig. 4B). GSH at this concentration had no effect on respiration (data not shown). When superoxide dismutase (SOD ; 1 U/ml) and catalase (1 U/ml) were present, they did not prevent the increase in state 4 respiration caused by DA and tyrosinase (Fig. 4B).

Figure 4.

Effect of enzymatic oxidation of DA on mitochondrial respiration. Mitochondria were incubated for 5 min in buffer or buffer containing DA (20 or 100 μM) + tyrosinase (Tyr ; 200 U/ml) with or without GSH (500 μM) or SOD (1 U/ml) + catalase (Cat ; 1 U/ml), and then respiration was measured using glutamate and malate as substrates. A : State 3 respiration. B : State 4 respiration. *Significantly different from control respiration (mean ± SEM ; n = 3-13).

FIG. 4.

Effects of DA oxidation products on mitochondrial permeability transition

Induction of permeability transition has been shown to lead to swelling of mitochondria (Gunter and Pfeiffer, 1990), which can be measured spectrophotometrically. Figure 5A shows the results of a representative swelling experiment. Exposure of brain mitochondria to control conditions (70 μM CaCl2) led to a small degree of mitochondrial swelling. Exposure to 70 μM CaCl2, followed by the addition of DA (100 μM), did not lead to an increase in the amount of mitochondrial swelling. However, exposure to the rapidly oxidizing catecholamine 6-OHDA (100 μM) significantly increased mitochondrial swelling. The results of these experiments, expressed as a change in absorbance, are quantified in Fig. 5B, demonstrating that 6-OHDA caused a threefold increase in mitochondrial swelling above control levels. CsA (850 nM), known to prevent the opening of the PTP in liver and heart mitochondria (Fournier et al., 1987 ; Crompton et al., 1988 ; Broekemeier et al., 1989), did not significantly prevent the swelling caused by 6-OHDA (Fig. 5B). CsA alone has no effect on swelling (data not shown).

Figure 5.

Effect of DA and 6-OHDA on swelling of brain mitochondria. A : Representative traces of the change in absorbance at 540 nm, indicative of swelling in brain mitochondria, are shown following exposure to CaCl2 (70 μM) alone (Control), CaCl2 + DA (100 μM), or CaCl2 + 6-OHDA (100 μM). CaCl2 was added after 0.5 min, and DA or 6-OHDA was added at 2 min. When utilized, CsA (850 nM) was present at the beginning of the experiment. Absorbance changes due to the autoxidation of 6-OHDA were subtracted using blanks containing only buffer and 6-OHDA. B : Quantification of swelling measured as the absolute change in absorbance from the time the inducer was added (2 min) to 10 min (mean ± SEM ; n = 3-6). *Significantly different from control values (p <0.05).

FIG. 5.

We also examined the effects of DA quinone production using tyrosinase (Fig. 6). We found that brain mitochondria exposed to CaCl2 (70 μM) followed by DA (100 μM) and tyrosinase (200 U/ml) led to a significant increase in mitochondrial swelling. In contrast to the findings with 6-OHDA, CsA (850 nM) was able to completely prevent the increase in swelling. Similar results were observed when a lower concentration of DA (20 μM) was utilized (Fig. 6). The addition of GSH (1 mM) also significantly reduced the amount of swelling induced by DA and tyrosinase to levels similar to controls (Fig. 6).

Figure 6.

Effect of tyrosinase (Tyr)-catalyzed oxidation of DA on swelling of brain mitochondria. A : Representative traces of swelling after exposure to CaCl2 (70 μM) alone (Control) or + DA and tyrosinase (200 U/ml), with or without the addition of GSH (1 mM) or CsA (850 nM). CaCl2 was added after 30 s, and then DA and tyrosinase were added at 2 min. When utilized, CsA and GSH were present at the beginning of the experiment. B : Quantification of swelling, measured as described in Fig. 5 (mean ± SEM ; n = 3-6). *Significantly diffeent from control values ; **significantly different from the same condition without CsA or GSH.

FIG. 6.

The degree of swelling induced by DA oxidation in brain mitochondria is much smaller than that classically observed in studies of inducers of the transition pore. Most compounds have been tested in liver mitochondria, and we have noted that brain mitochondria respond to a much smaller degree to classic inducers of the PTP than do liver mitochondria (S. B. Berman, S. C. Watson, and T. G. Hastings, unpublished data). Therefore, we also examined the effect of DA quinone production on liver mitochondria. We observed that exposure of liver mitochondria to CaCl2 (70 μM) followed by DA (100 μM) and tyrosinase (200 U/ml) resulted in large-amplitude swelling that was also prevented by the addition of CsA (Fig. 7). The change in absorbance induced in liver mitochondria by DA and tyrosinase was 11-fold higher than that observed in brain mitochondria (Figs. 6B and 7B) and was of similar magnitude to that observed for classic inducers of the PTP in liver mitochondria (e.g., Savage et al., 1991 ; Bernardi et al., 1992 ; S. B. Berman, S. C. Watson, and T. G. Hastings, unpublished data).

Figure 7.

Effect of tyrosinase (Tyr)-catalyzed oxidation of DA on swelling of liver mitochondria. A : Representative traces of swelling after exposure to CaCl2 (70 μM) alone (Control) or + DA and tyrosinase (200 U/ml), with or without CsA (850 nM). Experiments were conducted as described in Fig. 6, except utilizing mitochondria isolated from liver. B : Quantification of swelling, measured as described in Fig. 5 (mean ± SEM ; n = 3-6). *Significantly different from control values (p < 0.05) ; **significantly different from the same condition without CsA (p < 0.05). Note the different scale indicating absorbance change as compared with Fig. 6.

FIG. 7.


In this study, we provide evidence that both oxidation of the DA catechol ring and MAO-catalyzed DA oxidation can negatively affect the function of isolated brain mitochondria. We found that exposure of brain mitochondria to DA oxidation products altered both state 3 and state 4 mitochondrial respiration and caused mitochondrial swelling that may be indicative of the opening of the PTP.

Mitochondrial respiration

DA oxidation to DA quinone. We provide the first evidence that oxidation of DA to DA quinone can lead to changes in mitochondrial respiration. We found that exposure of mitochondria to DA quinone resulted in a large increase in the resting rate of mitochondrial respiration (state 4). Normally, the transport of electrons through the electron transport chain is coupled to the pumping of protons from the matrix across the inner mitochondrial membrane, which is generally impermeable to protons (Brand et al., 1994). This establishes a proton gradient across the membrane, and protons are driven back into the matrix through the ATP synthase, providing the energy to convert ADP to ATP. In this way, mitochondrial respiration and its resultant oxygen consumption are coupled to ATP synthesis. If mitochondrial respiration were completely coupled to ATP synthesis, resting state 4 respiration would be zero, as ATP synthesis is not occurring. In healthy mitochondria, a low rate of respiration persists under state 4 conditions, and this is though to be indicative of a low rate of proton leakage back across the inner membrane (Hafner et al., 1990). An increase in state 4 respiration therefore implies an increase in the proton leak across the membrane. After exposure to DA quinone, the rate of state 4 respiration increases nearly to the level of respiration measured in the presence of ADP (state 3). In other words, after exposure to DA quinone, respiration in the mitochondria proceeded without being coupled to ATP synthesis, an unproductive use of cellular energy.

It is possible that tyrosinase-catalyzed oxidation of DA produces not only the DA quinone but also O2•-, albeit at a much lower level (Tomita et al., 1984 ; however, see Koga et al., 1992). To begin to determine whether the quinone is important in the effect on mitochondrial respiration that was observed, we investigated protection by either GSH or SOD plus catalase. GSH can reduce the quinone to DA or utilize its sulfhydryl group to covalently bind to DA quinone (Tse et al., 1976), thus scavenging the quinone and preventing it from binding to mitochondrial proteins. GSH was largely able to prevent the effect of DA and tyrosinase, suggesting that DA quinone is involved in the increase in resting respiration. SOD and catalase, which would detoxify any O2 that might be formed, showed no protection against the effect caused by DA plus tyrosinase, further supporting the hypothesis that it is the DA quinone that is responsible for the uncoupling of respiration from ATP production.

Although the mechanism associated with the stimulation of resting respiration by DA quinone is not known, a similar effect on brain mitochondria was recently reported after exposure to the oxidant peroxynitrite (Brookes et al., 1998). This effect was prevented in the presence of the antioxidant Trolox. Although lipid peroxidation was suggested as a possible cause of the peroxynitrite effect, peroxynitrite, like DA quinone, can also modify sulfhydryl groups (Radi et al., 1991). Thus, it is also possible that covalent modification of sulfhydryl-containing proteins in the mitochondrial membrane contributes to the increase in proton permeability.

MAO-catalyzed oxidation of DA. We also found that MAO-catalyzed oxidation of DA, producing H2O2, led to an inhibition of active state 3 mitochondrial respiration. As both catalse and pargyline could prevent the inhibition, the production of H2O2 appears to be responsible for the effect. To determine where in the electron transport chain the inhibition was occurring, we utilized compounds that donate electrons directly to specific points along the chain, bypassing others. We found that inhibition of state 3 persisted when complex I was bypassed, suggesting that a process downstream of complex I is being affected. Furthermore, the inhibition disappeared when electrons were directly donated to cytochrome c just prior to complex IV, thus bypassing complex I and III. These findings suggest the possibility that the ubiquinone-complex III component of the electron transport chain is the target of the H2O2-induced inhibition. However, we cannot exclude the possibility that both complex I and II are inhibited by the H2O2. Both succinate dehydrogenase (complex II) and NADH dehydrogenase (complex I) have been shown to have thiol-dependent activity (Kenney, 1975 ; Benard and Balasubramanian, 1995) and thus may be susceptible to oxidation by H2O2.

This is the first study to examine the effect of DA on the oxygen consumption of well-coupled, healthy, intact mitochondria. Several previous studies examined only the effect of DA on complex I enzyme activity in tissue homogenates and reported conflicting results such as complete inhibition of complex I by 10 μM DA (Ben-Shachar et al., 1995), 25-50% inhibition with 1-100 mM DA (Przedborski et al., 1993), and 10% inhibition by 10 mM DA (Morikawa et al., 1996). The reasons for the discrepancies are not clear but may reflect different assay conditions. The two studies that began to investigate the role of DA oxidation in the enzyme inhibition noted prevention by antioxidants (Przedborski et al., 1993) and iron chelation (Ben-Shachar et al., 1995) but not by MAO inhibition (Ben-Shachar et al., 1995). All of these studies utilized only disrupted mitochondrial preparations to examine enzyme activities. However, studies utilizing intact, actively respiring mitochondria are likely to be more closely related to physiological circumstances, as the electron transport chain enzymes are normally coupled to ATP synthesis. In fact, studies have suggested that in nonsynaptic brain mitochondria, at least 70% inhibition of complex I is required before changes in state 3 respiration or ATP production are observed (Davey and Clark, 1996). Thus, the small inhibitory effect on respiration observed in our study may be indicative of larger changes in enzyme function. One other recent study utilized intact mitochondria but used extended incubation periods, and therefore less-coupled mitochondria, and utilized dye reduction rather than oxygen consumption as a measure of respiration, which does not allow examination of ATP synthesis-coupled respiration (Cohen et al., 1997). It also studied only one of the two physiologic pathways of DA oxidation, MAO-catalyzed oxidation of DA, and reported an inhibition of respiration dependent on this oxidation, similar to our results. With avoidance of some of these limitations in our methods, the results of the current study confirm this finding as well as demonstrate significant effects on respiration by DA oxidation to DA quinone.

Mitochondrial permeability transition

Opening of the mitochondrial PTP has been suggested to play a critical role in several forms of neuronal cell death, including excitotoxicity, ischemia, MPP+-induced toxicity, and hypoglycemia (Uchino et al., 1995 ; Nieminen et al., 1996 ; Packer et al., 1996 ; Schinder et al., 1996 ; White and Reynolds, 1996 ; Ouyang et al., 1997 ; Friberg et al., 1998). The increase in inner mitochondrial membrane permeability that is induced by PTP opening leads to mitochondrial membrane depolarization, release of small solutes and proteins, osmotic swelling, and a loss of oxidative phosphorylation (see Gunter and Pfeiffer, 1990 ; Bernardi, 1995). The PTP has been shown to contain critical thiol residues (Chernyak and Bernardi, 1996), and it has been induced by many oxidants including quinones (Gunter and Pfeiffer, 1990 ; Henry and Wallace, 1995 ; Henry et al., 1995). Therefore, we also examined effects of DA oxidation products on the PTP in mitochondria, and we conclude that DA quinone production resulted in the opening of the mitochondrial PTP. Mitochondrial swelling was induced in brain mitochondria after exposure to DA quinone that was completely prevented by the presence of the PTP inhibitor CsA, suggesting involvement of the PTP. GSH, which can inhibit DA quinone from reacting with sulfhydryl-containing mitochondrial proteins by covalently binding it as well as reducing it, was also able to prevent the swelling induced by DA quinone production.

Classically, most studies of the pore are performed in liver mitochondria, where a robust response has been well characterized (see Bernardi, 1995). To further investigate the effect of DA quinone production on classic pore characteristics, we also examined liver mitochondria. We found that enzymatic oxidation of DA produced large-amplitude swelling of liver mitochondria that was completely prevented by CsA, similar to that observed with other known pore inducers (e.g., Broekemeier et al., 1989 ; Bernardi et al., 1992). The magnitude of the swelling response in brain mitochondria was much smaller than that observed in liver (~10% of liver response) but is consistent with comparisons of PTP opening in brain and liver mitochondria after exposure to classic pore inducers such as calcium with phosphate (S. B. Berman, S. C. Watson, and T. G. Hastings, unpublished data).

The opening of the PTP induced by DA oxidation to DA quinone is interesting, given recent mechanistic studies of the transition pore. Evidence suggests that two sites exist on pore proteins that are important regulators of PTP function (Chernyak and Bernardi, 1996). One site contains vicinal thiols, which, when oxidized to disulfides, induce PTP opening. This site can be protected from oxidation by compounds that bind monothiols and prevent disulfide formation (Petronilli et al., 1994 ; Chernyak and Bernardi, 1996). One might expect that binding of DA quinone could protect similarly, binding directly to the monothiols. Benzoquinone has been suggested to inhibit PTP opening through this mechanism (Palmeria and Wallace, 1997). In contrast, there is also evidence that high concentrations of a monothiol-binding compound, N-ethylmaleimide, increased rather than decreased PTP opening (Petronilli et al., 1994), similar to the results with DA quinone.

A second important regulatory modulation involves the redox status of pyridine nucleotides. Oxidation of NADH and NADPH also increases pore opening, through an unknown mechanism that has been shown to be independent of the dithiol site (Chernyak and Bernardi, 1996). Oxidation of pyridines (both NADH and NADPH) can occur enzymatically through the cytosolic and mitochondrial enzyme DT-diaphorase, which reduces quinones via a two-electron reduction (Cadenas, 1995), and it has been shown that quinone substrates of DT-diaphorase can induce PTP opening (Chernyak and Bernardi, 1996). PTP opening by DA quinone could also be explained by this enzyme converting DA quinone to DA, oxidizing pyridine nucleotides in the process and increasing the probability of pore opening.

Availability of DA

For DA oxidation products to exert effects on mito-chondrial function, DA must be available to mitochondrial proteins. Although the majority of DA in DA neurons is stored in vesicles, much of the DA clearly has access to mitochondria, because MAO, the major metabolizing enzyme of DA, is located on the outer mitochondrial membrane (Greenawalt and Schnaitman, 1970). One could hypothesize that under conditions of increased availability of cytoplasmic DA or increased synthesis and metabolism of DA, the potential for DA oxidation-induced effects on mitochondria would increase. Such conditions are thought to exist both in PD, where there is an increase in DA turnover (Bernheimer et al., 1973), and following high doses of methamphetamine, resulting in the redistribution of DA from vesicular storage to the cytoplasm (Cubbells et al., 1994 ; Sulzer et al., 1995). In fact, DA oxidation products have been shown to be increased in the substantia nigra of postmortem brain tissue from parkinsonian patients (Fornstedt et al., 1989 ; Spencer et al., 1998) and in rat striatum following exposure to methamphetamine (LaVoie and Hastings, 1999). Thus, these conditions may lead to an increase in cytoplasmic DA and subsequent increase in DA oxidation products, resulting in eventual dysfunction of mitochondria. Although this study was performed in vitro with isolated mitochondria and relatively high concentrations of DA, it raises the possibility that DA oxidation-induced alterations in mitochondrial function could occur under pathological conditions.


The alterations in mitochondrial function due to DA oxidation have several potential implications for neuronal cell death and neurodegenerative disease. As mentioned previously, it has been reported that individuals with PD exhibit a deficiency in the activity of complex I of the electron transport chain. It is not entirely clear whether this is a deficiency in all cells (Parker et al., 1989 ; Shoffner et al., 1991 ; Martin et al., 1996 ; Sheehan et al., 1997) or is limited to the substantia nigra (Schapira et al., 1990b ; Mann et al., 1992), but the majority of evidence seems to point to a global deficiency. Therefore, the question arises as to the mechanism by which an underlying enzyme deficiency in all cells would lead to the specific loss of DA neurons. Our evidence suggests that one potential contributing factor may be the presence of DA. It is possible that an underlying deficiency, which alone does not cause cell death, is exacerbated by the presence of reactive DA metabolites. Other factors that have been implicated in PD, such as decreased antioxidant ability or increased iron (see Fahn and Cohen, 1992), also will contribute to increases in DA oxidation.

The effects of DA oxidation on mitochondrial function may also contribute to more acute neurotoxic events in which DA has been implicated. In methamphetamine toxicity, for example, DA is known to be important to the toxicity (Cubbels et al., 1994 ; Stephans and Yamamoto, 1994), and DA oxidation products have been shown to correlate with methamphetamine toxicity (La Voie and Hastings, 1999). Thus, DA-induced mitochondrial dysfunction may also play a role in this neurotoxicity.


This study was supported in part by USPHS grants DA09601 and NS19068 and USAMRMC grant 98292027.