INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The two major populations of dopamine (DA) neurons in midbrain are the substantia nigra pars compacta (SNc) and the ventral tegmental area (VTA). The DA cells of both regions share common electrophysiological characteristics (Grace and Bunney 1995), including somatodendritic release of DA (Geffen et al. 1976; Kalivas and Duffy 1991; Nieoullon et al. 1977; Rice et al. 1997). They differ in other aspects, however, including their major efferent projections: DA cells of the SNc project to the dorsal striatum (nigrostriatal system), whereas those of the VTA project to the nucleus accumbens and other mesolimbic structures (mesolimbic system) (Fallon and Moore 1978).

Another significant difference is that DA cells of the SNc degenerate in Parkinson's disease, whereas those in the VTA are spared (Fearnley and Lees 1991; Yamada et al. 1990). Several biochemical differences have been proposed to contribute to this, including higher expression of the DA transporter and lower levels of Ca²⁺-binding proteins in the SNc than in the VTA (Blanchard et al. 1994; German et al. 1992; Sanghera 1994). In addition, levels of the antioxidant enzymes catalase, glutathione (GSH) peroxidase, and Cu/Zn-superoxide dismutase are higher in VTA and nucleus accumbens than in SNc and dorsal striatum (Hung and Lee 1998; Trépanier et al. 1996). Although oxidative stress is an underlying factor in Parkinson's disease (Cohen et al. 1997; Olanow and Tatton 1999), few studies have addressed the functional consequences of these antioxidant differences.

We recently reported that DA release from synaptic terminals in dorsal striatum is reversibly inhibited by endogenous H₂O₂ generated during local stimulation (Chen et al. 2001). This suggests a normal regulatory process that could lead to oxidative stress if generation or regulation of H₂O₂ was disrupted. Because of the potential involvement of oxidative stress in DA neuron degeneration in the SNc, we have now investigated whether H₂O₂ might differentially modulate somatodendritic release in SNc and VTA. The influence of endogenous H₂O₂ in midbrain was indicated by changes in evoked DA release after inhibition of GSH peroxidase with mercaptosuccinate (MCS) (Sokolova et al. 2001; Ying et al. 1999). Companion studies assessed the effect of MCS on synaptic DA release in dorsal striatum and nucleus accumbens.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Slice preparation and voltammetric recording

Male Hartley guinea pigs (150-250 g) were deeply anesthetized with 40 mg/kg ip pentobarbital and decapitated. All animal handling procedures were in accordance with National Institutes of Health guidelines and were approved by the New York University School of Medicine Animal Care and Use Committee. Coronal striatal and midbrain slices (400 µM) were prepared and incubated as described previously (Chen et al. 2001; Rice et al. 1997); Ca²⁺ concentration was 1.5 mM. Evoked DA release was elicited using local pulse-train stimulation (10 Hz, 30 pulses) and monitored using carbon-fiber microelectrodes with fast-scan cyclic voltammetry (Chen and Rice 2001; Chen et al. 2001; Rice et al. 1997). The monitored substance was identified as DA under all conditions by the characteristic DA voltammograms recorded in each region (Fig. 1). All drugs and chemicals were from Sigma (St. Louis, MO).

View larger version (16K): [in this window] [in a new window]   Fig. 1. Voltammograms of dopamine (DA) during calibration (1 µM) and during the maximum increase in evoked [DA]o in striatum and midbrain under control conditions. Evoked responses were identified as DA by the characteristic peak potentials of voltammograms recorded in nucleus accumbens (NAcc), dorsal striatum (Str), substantia nigra pars compacta (SNc), and ventral tegmental area (VTA). Typical oxidation and reduction peak potentials were 0.60 and 0.25 V vs. Ag/AgCl, respectively.

Statistical analysis

Data are means ± SE (n = number of slices), illustrated as percent control. Differences between evoked [DA]_o in control slices and in the presence of H₂O₂ or MCS were assessed using one-way ANOVA followed by Kruskal-Wallis post hoc analysis. In SNc and VTA only single stimulus recording is possible (Rice et al. 1997) so that the responses in H₂O₂ or MCS were compared with pooled control data for each region. This also precluded drug washout studies in SNc and VTA. In dorsal striatum and nucleus accumbens, however, reproducible DA release could be evoked at 10-min intervals, which permits same-site control data for measurements taken during drug wash-in and washout (Chen et al. 2001).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In nucleus accumbens, exogenous H₂O₂ (15 min, 1.5 mM) (Chen et al. 2001) caused a reversible decrease in synaptic DA release evoked by pulse-train stimulation (Fig. 2). Control peak [DA]_o in accumbens was 2.62 ± 0.29 µM, which decreased by 35% in H₂O₂ (n = 7, P < 0.05). Similar H₂O₂ exposure also significantly inhibited somatodendritic DA release in midbrain (Fig. 2). In SNc, peak [DA]_o in control slices was 0.81 ± 0.08 µM (n = 13), which decreased by 58% in H₂O₂ (n = 8; P < 0.001). In VTA, control [DA]_o was 0.61 ± 0.06 µM (n = 16) with a 48% decrease in H₂O₂ (n = 10; P < 0.01). Importantly, 1.2 mM H₂O₂ did not decrease evoked [DA]_o in these regions, nor in dorsal striatum (not illustrated). The same sharp concentration dependence is seen in the effect of H₂O₂ on synaptic physiology in hippocampal slices (Avshalumov et al. 2000).

View larger version (28K): [in this window] [in a new window]   Fig. 2. Exogenous H2O2 reversibly inhibited evoked synaptic DA release in NAcc and somatodendritic release in SNc and VTA (see text for n values and statistics).

To determine whether DA release could be modulated by endogenous H₂O₂ produced during pulse-train stimulation, MCS (1 mM) was used to inhibit GSH peroxidase. Consistent with the expected effect of elevated H₂O₂ levels, MCS caused a reversible inhibition of DA terminal release in both dorsal striatum (37% decrease from control levels of 0.70 ± 0.05 µM; n = 7, P < 0.05) and nucleus accumbens (38% decrease; n = 7, P < 0.01; Fig. 3). In SNc, evoked somatodendritic DA release was also suppressed in MCS (30% decrease; P < 0.01, n = 5). In VTA, however, MCS had no effect (n = 6; Fig. 3).

View larger version (34K): [in this window] [in a new window]   Fig. 3. Inhibition of glutathione (GSH) peroxidase with mercaptosuccinate (MCS; 1 mM) caused a reversible inhibition of evoked DA release in Str and NAcc. Increased endogenous H2O2 in MCS also inhibited somatodendritic release in SNc but not VTA (see text for n values and statistics).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

These results show for the first time that somatodendritic DA release can be modulated by H₂O₂ in both SNc and VTA. Additionally, inhibition of DA release in the nucleus accumbens confirmed our previous observations of the effects of H₂O₂ on axon-terminal release. In our previous work, evoked [DA]_o in dorsal striatum was enhanced in the presence of the H₂O₂-degradative enzyme, catalase, which demonstrated the inhibitory action of endogenously generated H₂O₂ on DA release (Chen et al. 2001). Here, we used inhibition of GSH peroxidase to examine the effects of endogenously generated H₂O₂. Consistent with a normal modulatory role for H₂O₂, evoked [DA]_o was decreased by MCS in the SNc, as well as in dorsal striatum and nucleus accumbens.

Strikingly, MCS had no effect in the VTA. Because exogenous H₂O₂ suppressed DA release in this region (Fig. 2), the singular resistance of evoked DA release in the VTA to modulation of endogenous H₂O₂ levels could reflect either more limited H₂O₂ production or greater H₂O₂ regulation compared with other regions. Indeed, the data are consistent with the higher activity of peroxidase enzymes in VTA compared with SNc reported previously (Hung and Lee 1998; Trépanier et al. 1996). Thus higher catalase activity in the VTA might completely compensate for the loss of GSH peroxidase activity after inhibition by MCS. On the other hand, the nucleus accumbens has higher levels of GSH peroxidase and catalase than dorsal striatum, yet MCS was equally effective in both DA-terminal regions. The susceptibility of DA release in the accumbens to MCS shows that resistance to peroxidase inhibition in the VTA is not simply a specialized characteristic of mesolimbic DA cells (including their axon terminals). Moreover, this implies that regional differences in H₂O₂ production may also be a factor, with possibly greater H₂O₂ generation in the accumbens than dorsal striatum, but less in the VTA than SNc.

Sources and consequences of endogenous H₂O₂ generation

The major source of H₂O₂ generation in brain tissue is mitochondrial respiration. During the process of oxidative phosphorylation, a significant amount of O₂ consumed is diverted to form superoxide (O), which is the stoichiometric precursor of H₂O₂ (Chance et al. 1979). We have calculated previously that this process could transiently produce up to mM levels of H₂O₂ within the restricted compartment of a synapse (or dendrite), with absolute levels limited by the availability of local peroxidase enzymes (see Chen et al. 2001).

In addition to mitochondrial sources, H₂O₂ will be produced in DA cells and terminals by the DA metabolizing enzyme, monoamine oxidase (MAO) (Cohen et al. 1997). We suggested previously that MAO-dependent H₂O₂ production in dorsal striatum might provide feedback inhibition to decrease transmitter release (Chen et al. 2001). The present findings indicate that this could also occur in SNc and nucleus accumbens, although perhaps not in VTA. How H₂O₂ modulates DA release is not yet known. One attractive hypothesis, however, is that H₂O₂ may lead to hyperpolarization of DA cell membranes as shown previously in hippocampal pyramidal neurons (Seutin et al. 1995). Other physiological actions of H₂O₂ are only now being elucidated (for discussion, see Chen et al. 2001).

Conclusions

The present results indicate that somatodendritic DA release in the SNc, but not in the VTA, is susceptible to modulation by endogenously generated H₂O₂. This may represent a normal physiological process that could lead to pathophysiology if regulation were disrupted. As noted in the preceding text, Parkinson's disease, which involves selective degeneration of the nigrostriatal DA system, has been linked to oxidative stress. The present findings indicate a potentially critical difference in the management of reactive oxygen species between SNc and VTA that could contribute to this differential neuropathology.