- Top of page
- Materials and methods
The neurotoxic actions of methamphetamine (METH) may be mediated in part by reactive oxygen species (ROS). Methamphetamine administration leads to increases in ROS formation and lipid peroxidation in rodent brain; however, the extent to which proteins may be modified or whether affected brain regions exhibit similar elevations of lipid and protein oxidative markers have not been investigated. In this study we measured concentrations of TBARs, protein carbonyls and monoamines in various mouse brain regions at 4 h and 24 h after the last of four injections of METH (10 mg/kg/injection q 2 h). Substantial increases in TBARs and protein carbonyls were observed in the striatum and hippocampus but not the frontal cortex nor the cerebellum of METH-treated mice. Furthermore, lipid and protein oxidative markers were highly correlated within each brain region. In the hippocampus and striatum elevations in oxidative markers were significantly greater at 24 h than at 4 h. Monoamine levels were maximally reduced within 4 h (striatal dopamine [DA] by 95% and serotonin [5-HT] in striatum, cortex and hippocampus by 60–90%). These decrements persisted for 7 days after METH, indicating effects reflective of nerve terminal damage. Interestingly, NE was only transiently depleted in the brain regions investigated (hippocampus and cortex), suggesting a pharmacological and non-toxic action of METH on the noradrenergic nerve terminals. This study provides the first evidence for concurrent formation of lipid and protein markers of oxidative stress in several brain regions of mice that are severely affected by large neurotoxic doses of METH. Moreover, the differential time course for monoamine depletion and the elevations in oxidative markers indicate that the source of oxidative stress is not derived directly from DA or 5HT oxidation.
Several studies link the formation of reactive oxygen species (ROS) to the CNS neurotoxicity following METH administration. Increases in oxidized and reduced glutathione levels immediately following METH administration have been reported (Harold et al. 2000) suggesting a role for antioxidant mechanisms in reducing METH-induced toxicity. In support of this, METH-induced neurotoxicity is attenuated by concurrent treatment with antioxidants or free radical spin-trapping agents (De Vito and Wagner 1989; Cappon et al. 1996; Hirata et al. 1998; Yamamoto and Zhu 1998). Furthermore, METH toxicity is also reduced in mice that over-express the antioxidant enzyme CuZn superoxide dismutase (Cadet et al. 1994; Hirata et al. 1996). Additionally, salicylate trapping techniques demonstrate early elevations in striatal 2,3-dihydroxybenzoic acid (DHBA) indicative of acute increases in ROS formation following METH treatment (Giovanni et al. 1995; Fleckenstein et al. 1997; Yamamoto and Zhu 1998; Kita et al. 1999). These studies support the hypothesis that increased ROS formation may be a contributing factor to toxicity produced by METH.
Toxic doses of METH lead to brain lipid peroxidation and the formation of various oxidation products including malondialdehyde (MDA; Yamamoto and Zhu 1998). Most METH studies have reported elevations of lipid oxidation products using a coupled thiobarbituric acid spectrophotometric assay (thiobarbituric acid reactive substances or TBARs) which measures a variety of substances including MDA as well as contaminating carbonyls from proteins, nucleic acids and other lipid metabolites of non-ROS origin. However, despite its relatively lower assay sensitivity, these studies have reported elevated TBARs in striatum and hippocampus following METH treatment in rodents (Acikgoz et al. 1998; Kita et al. 2000; Wan et al. 2000), and these elevations persist for as long as seven days after the last dose (Wan et al. 2000). Furthermore, salicylate trapping studies performed at 7 days after METH demonstrate the persistence of ROS production for days following the initial toxic insult (Wan et al. 2000). These findings suggest that there might also be delayed mechanisms of ROS production, the latter being independent of acute METH-associated ROS formation.
Free radical formation, possibly resulting from the oxidation of DA or 5-HT (Ohmori et al. 1993; Cadet and Brannock 1998; Hirata et al. 1998; Wrona and Dryhurst 1998; Yamamoto and Zhu 1998; Berman and Hastings 1999; LaVoie and Hastings 1999) has been implicated in the neurotoxicity produced by METH. However, the extent to which these oxidation products contribute to neuronal damage in the monoaminergic innervated regions is unclear. If early onset ROS formation (i.e. which occurs within hours after METH treatment) derives from the direct oxidation of DA or 5-HT then it might be expected that elevations in oxidative markers would occur within the same time frame as the decreases in monoamine levels. Little is known about the time course for the formation of oxidative markers after METH treatment. DA and 5HT levels are rapidly reduced within only a few hours after METH exposure. Thus, one of the primary objectives of the study was to determine if increased formation of oxidative markers paralleled depletions in monoamine content in METH-treated mice.
It is also unclear to what extent ROS formation leads to oxidative damage to other macromolecules including proteins. Evidence that proteins are modified following METH exposure includes the formation of cysteinyl-DA adducts, formed by the nucleophilic attack of cysteinyl sulfur with DA-quinone intermediates (LaVoie and Hastings 1999). However, direct oxidation of protein side groups are not detected by this latter method. One reliable index of oxidative stress to proteins involves the measurement of protein carbonyl formation, resulting from the reaction of ROS with various protein side groups including lysine and arginine (Rice-Evans et al. 1991). Protein carbonyl measurements offer certain advantages over TBARs because of its lack of interference with other non-protein substances and are a more dependable measure of protein oxidative stress (Levine et al. 1990). Whether METH treatment affects protein carbonyls in the striatum or other brain regions is not known. Therefore, we have simultaneously measured TBARs and protein carbonyls to provide a more complete picture of the damage to proteins and lipids produced by ROS following METH treatment. We recently demonstrated a good correlation between elevations in protein carbonyls and TBARs in the hippocampus and striatum following exposure to an excitotoxin (Gluck et al. 2000). Thus, the second objective of the study was to determine whether protein carbonyls were elevated in various brain regions of mice after METH treatment and to assess how these increases compare with the formation of TBARs as a measure of lipid peroxidation within the same brain regions.
- Top of page
- Materials and methods
Various lines of evidence implicate ROS as mediators of neuronal damage following METH exposure. In this study we show that the administration of toxic doses of METH to mice produces brain region-selective and time-dependent elevations of protein carbonyls and TBARs, providing the first evidence that METH administration leads to ROS-mediated oxidative stress to proteins. Concentrations of TBARs and protein carbonyls were significantly increased in the hippocampus and striatum, marginally increased in the frontal cortex and unchanged in the cerebellum following METH administration. These results indicate that protein carbonyl measurements provide an additional reliable marker of ROS mediated oxidative stress, and in conjunction with TBARs measurements offers a complementary technique for detecting regional differences in oxidative markers following METH treatment. In affected brain regions, the magnitude and time course for protein carbonyl formation closely paralleled that of TBARs generation (i.e. greater concentrations were observed at 24 h than at 4 h after METH treatment). Furthermore, the highly significant correlation between levels of protein carbonyls and TBARs taken from the same sample in a given brain region demonstrates that protein and lipid oxidation is probably arising from similar mechanisms. This might be expected given the highly reactive and non-discriminatory nature of hydroxyl and superoxide radicals.
A cause and effect relationship between ROS production following METH administration (evidenced by increased elevations in oxidative markers) and subsequent nerve terminal damage (as measured by neurotransmitter content) remains unresolved. In order to examine whether early elevations in protein carbonyls and TBARs could accurately predict the extent of toxicity, we determined monoamine levels in a parallel group of euthanized mice 7 days after METH treatment. These measures indicated a profound toxicity to the dopaminergic and serotonergic terminals in the striatum as well as the serotonergic terminals in the hippocampus and frontal cortex. Unexpectedly, neurotransmitter loss following METH administration did not coincide with the elevations of oxidative markers measured at 4 h and 24 h post-METH treatment in the brain regions. For example, the striatum appeared to be the most severely damaged brain region based on the neurochemical deficits (70% 5-HT loss and > 95% DA loss at 7 days), but this region did not have the greatest relative increase in TBARs or protein carbonyls. Depletions in hippocampal 5-HT (38% loss) at 7 days were less than that observed in the striatum. Although hippocampal DA loss was not determined due to technical limitations, we have previously observed that DA content in the hippocampus of mice is more than 20-fold lower than in the striatum (unpublished observations). Thus, if catecholamine-associated oxidation is a primary source of ROS formation and considering the extensive amount of DA in the striatum, DA oxidation would be expected to contribute a much greater oxidative burden in the striatum than in the hippocampus. We would therefore have predicted that the relative increase in oxidative markers would have been higher in the striatum than other brain regions. However, levels of oxidative markers were similar in the striatum and the hippocampus.
A further example of the incongruity between oxidative markers and nerve terminal damage is illustrated by data obtained in the frontal cortex. TBARs and protein carbonyls were both substantially elevated in the striatum and hippocampus, and were significantly higher at 24 h than at 4 h post METH treatment. However, in the frontal cortex there were no substantial elevations in oxidative markers, although a small but statistically significant effect was observed with respect to treatment but not time. Neurochemically, however, it appears that all three regions appeared to have sustained substantial nerve terminal injury. The lack of substantial elevations in oxidative markers in the frontal cortex was unexpected because the 7 day measures of 5-HT and 5-HIAA indicated damage to serotonergic nerve terminals. Although it might be argued that the decrements in 5-HT and 5-HIAA in the frontal cortex observed at the 7 day time point are not indicative of toxicity but rather of pharmacological actions, this seems unlikely for several reasons. First, 5-HT and 5-HIAA concentrations in the frontal cortex at 7 days after METH treatment were significantly reduced and were similar to the reductions observed in the hippocampus, a brain region that showed substantial elevations in TBARs and protein carbonyls. Second, concentrations of NE in frontal cortex as well as the hippocampus were as profoundly depleted as 5-HT at 4 h (∼ 90% reductions), but unlike 5-HT, they had nearly recovered by 24 h and were completely recovered by 7 days. These findings are indicative of a pharmacological effect of METH on NE neurons and a toxic effect on 5-HT neurons. Furthermore, O'Callaghan and Miller (1994) have shown that decreases in 5-HT of the magnitude observed in our studies are associated with increases in glial acidic fibrillary protein and reflect neurotoxic damage in the cortex.
In summary, differential elevations of oxidative markers can occur in brain regions damaged by METH. However, our findings also suggest that a neurotoxic insult following METH treatment may not always be associated with measurable elevations in oxidative markers. Therefore, an important conclusion to draw from our studies is that increases in oxidative markers are not necessarily reflective or predictive of the degree of neurotransmitter loss and nerve terminal damage. The reasons for a mismatch between the magnitude of the oxidative stress for lipids and proteins produced by METH and the extent of neurotoxic damage that ensues in brain regions are unclear and warrant further investigation, but may arise from differences in antioxidant buffering capacities or differential susceptibilities to oxidative stress. Furthermore, our finding of serotonergic nerve terminal damage without significantly elevated oxidative markers in the cortex provides further support that METH may initiate additional pathophysiological cell injury pathways involving apoptosis or nitric oxide (Stumm et al. 1999; Deng et al. 1999; Imam et al. 1999, 2001; Kuhn and Geddes 2000).
The cellular source of ROS production in the METH-treated mice is not known. It is proposed that oxidation products derived from DA or 5-HT (for a review see Graham et al. 1978; Wrona and Dryhurst 1998; LaVoie and Hastings 1999) factor heavily into the toxicity associated with METH administration. A recent report by Wrona and Dryhurst (1998) summarized cumulative data and highlighted the potential neurotoxic role for a group of serotonin metabolites arising from superoxide-mediated oxidation. Dopamine quinones have also been implicated in METH toxicity (LaVoie and Hastings 1999). Exposure of neurons to METH leads to disruption of vesicular monoamine stores, increased amounts of neurotransmitter in the cytosol and enhanced release into the extracellular space (Sulzer et al. 1995). These adverse events are thought to initiate the oxidative stress that leads to oxidative damage. However, monoamines are rapidly depleted following initiation of METH administration within only a few hours after the last METH injection; see (LaVoie and Hastings 1999) and the present findings. In the striatum, DA and 5HT are depleted by over 75% as early as 1 h after the last injection of METH (unpublished observations). Oxidative markers were not significantly or only modestly elevated in the hippocampus and striatum, respectively, at 4 h after METH whereas substantially greater amounts were observed at 24 h. These data show that monoamine depletion clearly precedes evidence of ROS-induced damage to cellular proteins or lipids. In contrast, protein cysteinyl-DA levels peak as early as 2 h after the last injection of METH, suggesting a rapid but short-lived formation of DA-quinone attack on proteins (LaVoie and Hastings 1999).
Based on these observations, it appears that the formation of ROS may depend initially on oxidation products derived from the monoamines but that continued ROS formation is independent of monoamine oxidation. If ongoing release of monoaminergic neurotransmitters does not account for the increase in oxidative markers at 24 h then additional pathophysiologic mechanisms must be operative. Continued formation of ROS independent of catechol or indole monoamine oxidation could occur with compromised mitochondrial function and aberrant electron transport function leading to increased production of superoxide and decreased ATP production (Burrows et al. 2000a; Lotharius and O'Malley 2000; Nakamura et al. 2000). There are several findings to suggest that mitochondrial function and energy production may be significantly decreased following METH treatment (Chan et al. 1994; Li and Dryhurst 1997; Berman and Hastings 1999). Additionally, coadministration of METH with mitochondrial inhibitors exacerbates toxicity in mice and rats (Albers and Sonsalla 1995; Bowyer et al. 1996; Burrows et al. 2000b). In a recent study, Yamamoto and colleagues reported a rapid but transient decrease in cytochrome oxidase staining in the striatum of METH treated rats (Burrows et al. 2000a), which supports related studies demonstrating decreases in striatal ATP in METH-treated mice (Chan et al. 1994). These studies are consistent with a METH-induced impairment of mitochondrial function. In agreement with these findings, our laboratory has recently observed that methamphetamine is equipotent as 1-methyl-4-phenylpyridinium ion (MPP+) in inhibiting rotenone sensitive NADH oxidase and NADH-ubiquinone reductase activities in submitochondrial particles isolated from pig heart (unpublished data), although its relevance to effects on intact mitochondria remain unclear. Nonetheless, it is possible that the delayed increase in oxidative stress seen in the hippocampus and striatum in our study may reflect direct METH-induced mitochondrial dysfunction. Future studies will need to be performed to obtain a more thorough analysis of the time-course formation of oxidative markers following METH administration as well as further clarifying any inhibitory effect of METH on mitochondrial function and the electron transport chain.