Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The abuse of psychostimulants, such as methamphetamine (METH), is a serious worldwide problem. Exposure to high doses of such stimulants results in partial damage to central monoamine systems, including the dopamine (DA) input to the striatum (Hotchkiss and Gibb, 1980; Wilson et al., 1996). In rodents, these depletions persist for at least 6 months, whereas effects in nonhuman primates reportedly are still present after 4 years (Bittner et al., 1981; Woolverton et al., 1989), suggesting that such exposure to METH may have long-lasting consequences on central nervous system function and, ultimately, behavior. In fact, administration of amphetamine or METH is associated with long-term cognitive and subtle motor deficits, an effect also caused by a partial loss of the DA innervation of the striatum (Walsh and Wagner, 1992; Fornaguera et al., 1993; Kirik et al., 1998). For example, chronic abusers of amphetamines display deficits in decision-making processes, and METH-treated animals exhibit deficits in reaction-time tasks (Richards et al., 1993; Rogers et al., 1999). Despite the persistence of these monoamine and behavioral deficits, the long-term impact of psychostimulant-induced damage to central monoamine systems on basal ganglia function and the alterations in central nervous system function underlying the cognitive and motor deficits are unknown.

Although the consequences of large DA depletions (>90%) on basal ganglia function have been well characterized, less is known about the effects of smaller depletions, such as those produced by METH. On the one hand, near total lesions of the dopamine innervation of the striatum lead to alterations in the activity of both principal striatal efferent pathways, as reflected by an increase in preproenkephalin (ENK) and zif 268 mRNA expression in striatopallidal neurons, and a decrease in preprotachykinin (SP) and zif 268 mRNA expression in striatonigral neurons (Gerfen et al., 1991, 1995). Furthermore, cytochrome oxidase (CO) histochemistry reveals an increase in neuronal activity in the subthalamic nucleus (STN), entopeduncular nucleus (EPN), and substantia nigra pars reticulata (SNr; Vila et al., 1996) after similar extensive lesions to DA neurons. On the other hand, partial loss of DA (<80%) induced by the neurotoxin 6-hydroxydopamine decreases SP mRNA expression, but not ENK mRNA expression (Nisenbaum et al., 1996). These data suggest that partial loss of the DA innervation of the striatum preferentially alters striatonigral neuron function.

The purpose of the present study therefore was to assess the long-term consequences of METH-induced DA depletion on basal ganglia function. We hypothesized that the partial DA depletion induced by a neurotoxic regimen of METH would lead to alterations in the function of the striatonigral pathway, while having little impact on the striatopallidal pathway. We report a selective decrease in striatal SP gene expression using in situ hybridization histochemistry, as well as alterations in neuronal activity in the EPN and SNr as measured by CO histochemistry 3 weeks after the administration of multiple high doses of METH. Additionally, rats similarly treated with METH exhibited behavioral deficits in a sequential motor learning task, indicating that exposure to a neurotoxic regimen of METH results not only in changes in central monoamine systems but also in the function of basal ganglia pathways and basal ganglia-mediated behaviors.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Animals. Male Sprague-Dawley rats (230-330 g; Simonsen Laboratories, Gilroy, CA) were used in all experiments. Rats were housed in groups in hanging wire cages in a room controlled for temperature and lighting, with free access to food and water. All animal care and experimental manipulations were approved by the Institutional Animal Care and Use Committee of the University of Utah, Salt Lake City, and were in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals.

Drugs. (±)-Methamphetamine hydrochloride was kindly provided by the National Institute on Drug Abuse (Rockville, MD). Drug doses were calculated as the free base.

Pharmacological Manipulations. The night before the experiment, rats were rehoused in plastic tubs (8-10 rats per tub) and weighed. The next morning, the rats received four injections of either saline (control) or METH (5.0, 7.5, or 10 mg/kg s.c.), at 2-h intervals. Rats remained in their tubs until the next day, when they were returned to their home cages. Rats were sacrificed 3 weeks after METH treatment by exposure to CO₂ (1 min). Rats were then decapitated, and the brains were rapidly removed and frozen in isopentane chilled on dry ice.

In Situ Hybridization Histochemistry. Frozen brains from control and METH-treated rats were cut into 12-µm sections in a cryostat (Cryocut 1800; Cambridge Instruments, Heidelberg, Germany). Sections were thaw-mounted onto gelatin-chrome alum-subbed slides and stored at 20°C. Once all brains from an experiment had been sectioned, slides were postfixed in 4% paraformaldehyde/0.9% NaCl, acetylated in fresh 0.25% acetic anhydride in 0.1 M triethanolamine/0.9%NaCl (pH 8.0), dehydrated in an ascending series of alcohols, delipidated in chloroform, and then rehydrated in a descending series of alcohols. Slides were air dried and stored at 20°C.

CO Histochemistry. Cytochrome oxidase histochemistry was performed on slide-mounted brain sections containing basal ganglia output nuclei using modifications to Wong-Riley (1979). Briefly, 20-µm sections were stored at 20°C until use. Sections were incubated in 300 ml of 0.1 M PBS (pH 7.4) containing 180 mg of 3,3-diaminobenzidine tetrachloride (DAB; Sigma, St. Louis, MO) and 45 mg of cytochrome c (from horse; Sigma) for 2.5 h at 37°C. Slides were then rinsed 3 × 10 min in 0.1 M PBS, air dried, and coverslipped with Permount (Fisher Scientific, Springfield, NJ). Sections were taken at the level of 1) caudal striatum/globus pallidus (GP; 1.3 mm posterior to Bregma), 2) EPN (2.3 mm posterior), 3) STN (3.8 mm posterior), and 4) SNr (5.3 mm posterior). The assay was validated by showing that increases in CO content corresponded with a linear increase in signal intensity (r² = 0.94; data not shown). To do so, whole brains were removed and homogenized. Half of the homogenate (by weight) was microwaved for 30 s to eliminate any endogenous enzyme activity. Brain homogenates were then mixed in varying ratios of microwave/homogenate to intact/homogenate (4:0, 3:1, 2:2, 1:3, 0:4). These mixtures were frozen and then sectioned at 20 µm and processed as described above.

TH Immunohistochemistry. TH immunohistochemistry was performed on nonperfused, 12-µm-thick sections using the peroxidase anti-peroxidase method. Slides were washed in 0.1 M PBS (pH 7.4), fixed in 4% paraformaldehyde/0.9% NaCl for 10 min, rinsed in PBS, and preincubated in 2% H₂O₂ in PBS for 20 min. After drying, slides were incubated overnight in mouse anti-TH antibody (1:500 dilution; DiaSorin, Stillwater, MN) in PBS, 0.3% Triton X-100, and 3% normal horse serum at 4°C. Slides were then washed in PBS and incubated in horse anti-mouse IgG (1:150 dilution; Vector Laboratories, Burlingame, CA) in PBS and 0.3% Triton X-100 for 3 h at room temperature. After rinsing in PBS, slides were incubated 1 h at room temperature in mouse peroxidase anti-peroxidase (dilution 1:500; Sigma) in PBS and 0.3% Triton X-100. Following a rinse in Tris-buffered saline (TBS; pH 7.6), slides were preincubated in 0.1% DAB in TBS, and incubated in 0.1% DAB/0.005% H₂O₂/TBS. Slides were then dehydrated in an ascending series of ethanol, cleared in xylene, and coverslipped.

DA Tissue Content. DA content in striatal tissue was determined in tissue punches collected during sectioning of the striatum for other assays. A blunt-tip 18-gauge needle was used to collect 1-mm³ punches from both the medial and lateral striatum (+0.3 mm anterior to bregma). Tissue punches were sonicated in tissue buffer [0.05 M sodium phosphate/0.03 M citric acid buffer with 25% methanol (v/v), pH 2.5] and centrifuged. Twenty microliters of supernatant was injected onto a high pressure liquid chromatography system coupled to an electrochemical detector (EOx = +0.6 V) for separation and quantitation of DA levels as previously described (Chapin et al., 1986). To adjust for variability in the size of the tissue punches, all values were expressed per milligram of protein. Protein content was determined with the Lowry protein assay.

Sequential Motor Learning Task. Ten days after treatment with either saline (control) or METH (10 mg/kg; four doses at 2-h intervals), rats were transported to new housing, and food intake was restricted as described above. Animals were handled daily for 3 days by the experimenter and then were introduced to the testing apparatus, consisting of a radial eight-arm maze. A food well was at the end of each arm, in which rewards (Froot Loops, Kellogg Company, Battle Creek, MI) were placed. Rats were allowed free access to all arms for a minimum of 10 min for five consecutive days to habituate them to the environment. Three weeks after METH administration, the behavioral trials began. For each rat a unique, fixed sequence of maze arm openings was established. The rat was placed in the middle of the maze with all doors leading out from the center closed. The first door of the sequence was opened and the rat was allowed to move to the food well to collect its reward. Upon entrance of the rat into the first arm, the door of the arm next in the sequence was opened. The time taken for the rat to retrieve the food reward after crossing the plane of the door of the previous arm was recorded for arms 2 to 6 and summated to give the time to complete each trial. Animals completed four trials per day for five successive days, using the same sequence of door openings. They then completed 2 days (four trials per day) of trials in which the sequences of the door openings were randomized. Two additional days (four trials per day) were then completed in which the sequence of door openings was the previously used fixed sequence.

Data Analysis. Film autoradiograms and slides on which in situ hybridization histochemistry and CO histochemistry were performed were analyzed using the image analysis program Image (Wayne Rasband, National Institutes of Health, Bethesda, MD), to obtain average density (gray) values of given areas. The linearity of the video camera and video capture card to increasing signal intensity was determined by measuring the average gray values of signals of known optical density from a photographic step tablet (Eastman Kodak, Rochester, NY). The intensity of the illuminating light was adjusted until measurements obtained from film autoradiograms and CO-stained sections were within the linear parameters for the system. Lighting and camera conditions were kept constant throughout the capture and measurement process for each group of control and treated sections that were processed and neurochemically evaluated in parallel. For in situ hybridization histochemistry, measurements were made over the medial, lateral, and whole striatum at the rostral (1.7 mm anterior to bregma), middle 1 (0.5 mm anterior), middle 2 (0.4 posterior), and caudal (1.3 posterior) levels, as well as the substantia nigra pars compacta (SNc) and the ventral tegmental area. To account for background labeling, the average gray value of the white matter above the region of interest was subtracted from the average gray value of the regions of interest. For CO histochemistry analysis, the EPN was magnified 32× and only labeled neuropil, excluding white matter, was outlined and measured. Background labeling was determined by measuring the surrounding white matter. Images of the STN and SNr were magnified 32 and 24×, respectively, outlined, and measured.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

DA Depletion

Administration of multiple high doses of METH resulted in a dose-dependent decrease in the DA innervation of the striatum as evidenced by decreased TH mRNA in the SNc and decreased TH-immunoreactivity and DA tissue content in the striatum. In situ hybridization histochemistry revealed an approximate 20% decrease in TH mRNA in the SNc 3 weeks after the administration of the highest METH dose (p < 0.03; Table 1; Fig. 1A). Analysis of the ventral tegmental area showed no decrease in TH mRNA (data not shown). Immunohistochemical analysis of TH protein in striatum revealed a decrease in TH-like immunoreactivity at all levels of the striatum (Fig. 1B). METH administration decreased DA tissue content in a dose-dependent manner in both the medial and lateral striatum (Table 1; p < 0.03 medial; p < 0.0001 lateral). The DA depletion was significantly greater in the lateral striatum than in the medial striatum at the highest dose (p < 0.03).

View larger version (111K): [in this window] [in a new window]   Fig. 1.   A, representative in situ hybridization autoradiograms of TH mRNA in the substantia nigra pars compacta of a saline-injected animal (control) compared with a METH-treated animal (4 × 10 mg/kg) 3 weeks after treatment. B, representative immunohistochemical sections of TH-like immunoreactivity in the middle striatum of a saline-injected animal (control) compared with a METH-treated animal (4 × 10 mg/kg) 3 weeks after treatment.

Changes in Neurochemistry of the Basal Ganglia

Neuropeptide Gene Expression. The DA depletion present 3 weeks post-METH treatment was associated with a decrease in SP mRNA levels, which was significant throughout all levels of the striatum (rostral level p < 0.003; middle 1 level p < 0.0005; middle 2 level p < 0.008; caudal level p < 0.002; Fig. 2, A-D). Only the high dose of METH resulted in a significant reduction of SP mRNA at the rostral level (Fig. 2A), whereas both the 7.5- and the 10-mg/kg METH doses significantly decreased SP mRNA in middle 1 and middle 2 sections (Figs. 2, B and C, and 3). Finally, in the caudal region of the striatum, SP mRNA was significantly decreased by all doses of METH (Fig. 2D). The decrease of approximately 25% in SP mRNA induced by METH was not significantly different in the medial versus lateral striatum. Although prior treatment with a neurotoxic regimen of METH decreased SP mRNA, it had no significant effect on the mRNAs for DYN, ENK, and NT in any region of the striatum (data not shown).

View larger version (28K): [in this window] [in a new window]   Fig. 2.   Quantitative changes in the amount of substance P mRNA in the rostral (A), middle 1 (B), middle 2 (C), and caudal striatum (D) 3 weeks following administration of four doses of METH at 2-h intervals. Values are mean gray values (arbitrary units; ±S.E.M.; n = 10 per group) obtained from densitometric analysis of film autoradiograms. *, significantly different from control (0 METH), p < 0.05.

View larger version (109K): [in this window] [in a new window]   Fig. 3.   In situ hybridization film autoradiogram of substance P mRNA in the middle 1 section of a saline-injected animal (control) and a methamphetamine-treated animal (4 × 10 mg/kg) 3 weeks after treatment.

CO Histochemistry. Administration of a neurotoxic regimen of METH 3 weeks earlier produced an increase in CO activity in the EPN (Figs. 4A and 5) and SNr (Fig. 4B). The increase was significant in the SNr for all doses of METH (p < 0.0005), and in the EPN for the two highest doses (p < 0.0001). No changes in CO activity were seen in the caudal striatum, GP, or STN (data not shown).

View larger version (15K): [in this window] [in a new window]   Fig. 4.   Quantitative changes in the amount of CO activity in the entopeduncular nucleus (A) and substantia nigra pars reticulata (B) 3 weeks after administration of four doses of METH (5-10 mg/kg). Values are mean gray values (±S.E.M.; n = 8-10 per group) obtained from densitometric analysis of slides stained for CO activity. *, significantly different from control (0 METH), p < 0.05.

View larger version (96K): [in this window] [in a new window]   Fig. 5.   Representative images of cytochrome oxidase histochemical staining in sections at the level of the entopeduncular nucleus from a saline-injected animal (control) and a methamphetamine-treated animal (4 × 10 mg/kg) 3 weeks after treatment.

Long-Term Effects of METH Administration on a Sequential Learning Task

A separate group of rats treated with multiple high doses of METH (4 × 10 mg/kg) had a 43 to 55% loss of striatal DA tissue content when the striata were analyzed 6 weeks after the injections of METH. These rats were impaired on a radial maze, sequential learning task when tested 3 weeks after the injections of METH. Two-way ANOVA revealed a significant interaction (treatment × trial; p < 0.0005). Post hoc analysis revealed that METH-treated rats took significantly longer to run the fixed sequence at almost all time points (Fig. 6). However, when running the random sequences, METH-treated rats completed the trials as fast or faster than their saline-injected counterparts (random A and B; p < 0.05). Upon returning to the fixed sequence, METH-treated rats were once again significantly slower than the controls (fixed 6 and 7; p < 0.01).

View larger version (20K): [in this window] [in a new window]   Fig. 6.   Time required for rats to complete an eight-arm radial maze in which the sequence in which the doors were opened was fixed (fixed 1-7) or random (random A and B). Each rat was assigned a unique fixed sequence of door openings. The time to complete the maze was averaged over four daily trials for each rat. Values shown are the median times per day (±S.E.M.) for control (n = 8) or METH-treated (4 × 10 mg/kg; n = 9) rats. Fixed 1 to fixed 5 and fixed 6 to fixed 7 indicate the days on which rats ran the maze with the doors being opened in the fixed sequence established for that rat. Random A and random B indicate the days on which rats ran the maze with the doors being opened in a random sequence (different on every trial). *, significantly different from control on the same trial, p < 0.05.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Although neurochemical and behavioral changes associated with near-total losses of DA in striatum have been well characterized, little is known about the long-term consequences on basal ganglia function of the partial monoamine loss associated with METH-induced neurotoxicity. The present data show that 3 weeks after administration of a neurotoxic regimen of METH, SP mRNA in striatum was decreased and CO activity in the EPN and the SNr was increased. However, ENK mRNA in the striatum and neuronal activity in the GP and STN were unaltered. These data suggest that such exposure to METH results in a selective decrease in striatonigral pathway function.

Numerous studies indicate that the DA innervation of striatum is compromised following administration of high doses of METH (Hotchkiss and Gibb, 1980; Wagner et al., 1980). In the present study, dose-dependent decreases in DA tissue content and TH immunoreactivity in striatum were seen 3 weeks after administration of multiple doses of METH. At present, the significance of the decrease in TH mRNA levels in the SNc is not clear. Previous studies suggest that the number of cells in the SNc is not decreased after neurotoxic regimens of METH (Ricaurte, et al., 1982; O'Callaghan and Miller, 1994), although stereology-based cell counts have not been performed to conclusively address this issue. Other studies, however, note decreases in dopamine transporter mRNA and the number of TH-immunoreactive neurons in the SNc (Sonsalla et al., 1996; Hirata and Cadet, 1997). Thus, it is currently not known whether the decrease in TH mRNA represents decreased TH synthesis in nigral dopamine neurons or actual loss of cell bodies. Despite this uncertainty, it is clear that multiple high doses of METH produce persistent deficits in the DA innervation of striatum.

Although the toxic effects of METH on monoamine systems have long been known, the postsynaptic consequences of this loss on basal ganglia function have not been well studied. We found that 3 weeks following administration of a neurotoxic regimen of METH, SP mRNA levels in striatum were significantly decreased in a dose-related manner. A possible mechanism underlying the change in SP mRNA is its localization in D1 DA receptor-containing striatonigral neurons (Gerfen et al., 1990; Hersch et al., 1995). Because D1 receptors have a lower affinity for DA than do D2 receptors (Seeman and Grigoriadis, 1985; Dearry et al., 1990), striatonigral neurons may be more sensitive to partial loss of DA innervation than striatopallidal projections, which are dominated by D2 receptors. This possibility is supported by the observation that the decrease in SP mRNA induced by near total DA depletions is reversed by intrastriatal grafts only in regions directly innervated by the graft (Cenci et al., 1993). Although DYN is coexpressed in striatonigral neurons, its basal expression is often not affected by even near-total decreases in DA (Gerfen et al., 1991). We also did not observe any persistent change in DYN mRNA after METH treatment. On the other hand, the expression of ENK and NT mRNAs appears to be altered only by large DA depletions (Gerfen et al., 1991; Nisenbaum et al., 1996; Hanson and Keefe, 1999), consistent with the lack of change observed in the present study. These selective changes in SP mRNA, in conjunction with changes in CO activity only in the terminal fields of striatonigral neurons, suggest that METH-induced neurotoxicity produces selective alterations in striatonigral efferent neuron function.

The observed changes in SP expression also may be a consequence of METH-induced decreases in striatal 5-HT (Hotchkiss and Gibb, 1980). Although we did not measure 5-HT concentrations in the striata of the rats included in the present neurochemical analyses, determination of 5-HT content in striata of rats treated with 4 × 10 mg/kg subcutaneous doses of METH, including those used in the present behavioral experiments, revealed a 30 to 50% loss of 5-HT (data not shown). Serotonin systems have been shown to regulate SP mRNA expression in striatum through 5-HT2a/2c receptors, and lesions of 5-HT with 5,7-dihydroxytryptamine result in decreased SP mRNA expression (Walker et al., 1996; Gresch and Walker, 1999). In addition, 5-HT innervation of striatum and 5-HT2a receptor expression are greatest in caudal striatum (Ternaux et al., 1977; Pompeiano et al., 1994). The changes in striatonigral neuron function observed in this study in response to 5 mg/kg METH, which does not significantly alter striatal DA, might thus result from 5-HT depletion, although the dose-dependent effects of METH on 5-HT remain to be determined. However, it should be noted that depletion of 5-HT in striatum also decreases Enk mRNA levels (Walker et al., 1996), a change not seen in the present study. Thus, the profile of changes observed in the present study parallels better the changes resulting from partial DA, rather than partial 5-HT, depletions.

A decrease in striatonigral, but not striatopallidal, neuron function by METH-induced neurotoxicity is further supported by the increases in cytochrome oxidase activity in the EPN and SNr, as measured by CO histochemistry. CO is a reliable marker for long-term changes in neuronal activity (Wong-Riley et al., 1998). Striatonigral neurons send their primary axonal projections to the EPN and SNr (Kawaguchi et al., 1990). Decreased function of these neurons resulting from reduced D1 receptor activation consequent to partial loss of DA therefore would lead to disinhibition of EPN and SNr neurons, and increased CO activity. Near-total DA depletions, on the other hand, alter both striatopallidal and striatonigral neuron function, resulting in increased CO activity in the internal segment of the globus pallidus and the STN, and increases in CO subunit I mRNA in the internal segment of the globus pallidus, SNr, and STN in nonhuman primates (but see Porter et al., 1994; Vila et al., 1996, 1997). Again, we cannot exclude the possibility that METH-induced damage to striatal 5-HT systems also contributed to the observed changes in CO activity. Regardless of the relative roles of DA and 5-HT loss in these effects, the alterations in activity of neurons receiving projections from striatonigral neurons as a consequence of METH-induced neurotoxicity, and the lack of changes in nuclei modulated by striatopallidal efferents, support the conclusion that long-lasting postsynaptic changes are limited to the striatonigral pathway after high-dose METH administration.

The pre- and postsynaptic changes produced by METH in the current study were associated with impairment of the rats' performance on a sequential motor learning task 3 weeks following drug administration. Although it appears that METH-treated rats were able to learn the fixed sequence, as indicated by the similar decrease in running time over fixed trials 1 to 5 for both control and METH-treated rats, they were significantly slower than saline-injected rats at a number of points. This could be attributed to a motor deficit secondary to loss of DA. However, METH-treated animals completed the trials of random sequences (random trials A and B) at equal or significantly faster rates than did controls, suggesting that the slower times on fixed sequences were not due to a general performance deficit. The observation that the running times for fixed sequences after the random trials were again slower in the METH-treated animals confirms the specificity of this deficit and suggests the presence of a METH-induced deficit in sequential motor learning.

The neurochemical changes responsible for the cognitive impairment observed after exposure to METH and other amphetamines remain to be determined. Cognitive deficits have been noted in primates with partial DA depletions induced by 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (Fernandez-Ruiz et al., 1995; Schneider and Pope-Coleman, 1995). Furthermore, similar deficits in sequential motor learning have recently been reported in primates with 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine-induced depletion of DA in the caudate-putamen (Matsumoto et al., 1999), suggesting that this deficit may be a consequence of the METH-induced decrease in striatal DA. Serotonin deficiency, however, has been implicated in the cognitive deficits seen in chronic amphetamine abusers (Rogers et al., 1999). Finally, the present data suggest that SP may also play a critical behavioral role, because exposure to a neurotoxic regimen of METH was associated with a decrease in striatal SP mRNA. Although the actions of SP on the central nervous system remain unclear, this peptide is involved in determining functional outcome from DA-depleting brain lesions (Nikolaus et al., 1997). Whether manipulations or these neurochemical systems will restore sequential learning on this task remains to be determined.

In summary, neurotoxic doses of METH bring about prolonged changes in striatal monoamines. Our studies show that this same regimen of METH results in a selective reduction in the expression of SP mRNA in striatonigral neurons of the striatum, and increases in CO activity in the EPN and SNr, the terminal fields of striatonigral neurons. However, no persistent changes in neurochemical markers of striatopallidal neurons were observed after METH treatment. These results thus suggest that METH-induced neurotoxicity is associated with long-term decreases in the function of striatonigral efferent neurons and therefore persistent alterations in the function of the basal ganglia. Whether these deficits continue for longer periods of time, and the extent to which these deficits contribute to the persistent behavioral deficits observed, remain to be determined.