METHAMPHETAMINE-INDUCED CELL DEATH: SELECTIVE VULNERABILITY IN NEURONAL SUBPOPULATIONS OF THE STRIATUM IN MICE

Abstract

Methamphetamine (METH) is an illicit and potent psychostimulant, which acts as an indirect dopamine agonist. In the striatum, METH has been shown to cause long lasting neurotoxic damage to dopaminergic nerve terminals and recently, the degeneration and death of striatal cells. The present study was undertaken to identify the type of striatal neurons that undergo apoptosis after METH. Male mice received a single high dose of METH (30 mg/kg, i.p.) and were killed 24 h later. To demonstrate that METH induces apoptosis in neurons, we combined terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling (TUNEL) staining with immunohistofluorescence for the neuronal marker neuron-specific nuclear protein (NeuN). Staining for TUNEL and NeuN was colocalized throughout the striatum. METH induces apoptosis in approximately 25% of striatal neurons. Cell counts of TUNEL-positive neurons in the dorsomedial, ventromedial, dorsolateral and ventrolateral quadrants of the striatum did not reveal anatomical preference. The type of striatal neuron undergoing cell death was determined by combining TUNEL with immunohistofluorescence for selective markers of striatal neurons: dopamine- and cAMP-regulated phosphoprotein, of apparent M_r 32,000, parvalbumin, choline acetyltransferase and somatostatin (SST). METH induces apoptosis in approximately 21% of dopamine- and cAMP-regulated phosphoprotein, of apparent M_r 32,000-positive neurons (projection neurons), 45% of GABA-parvalbumin-positive neurons in the dorsal striatum, and 29% of cholinergic neurons in the dorsal–medial striatum. In contrast, the SST-positive interneurons were refractory to METH-induced apoptosis. Finally, the amount of cell loss determined with Nissl staining correlated with the amount of TUNEL staining in the striatum of METH-treated animals. In conclusion, some of the striatal projection neurons and the GABA-parvalbumin and cholinergic interneurons were removed by apoptosis in the aftermath of METH. This imbalance in the populations of striatal neurons may lead to functional abnormalities in the output and processing of neural information in this part of the brain.

Methamphetamine (METH) is a potent and addictive psychostimulant. The neurotoxic effects of this substituted amphetamine are associated with its ability to induce an overflow of dopamine in the synapse by displacing vesicular dopamine stores (Raiteri et al., 1979; Liang and Rutledge, 1982a,b; Schmidt et al., 1985; Sulzer et al., 1995). Displaced dopamine molecules and its metabolites can be readily oxidized to reactive quinones and semiquinones further generating reactive oxygen (Cadet and Brannock, 1998) and nitrogen radicals (Lipton and Rosenberg, 1994; Imam et al., 1999) that affect a wide range of modifications of proteins, sugars, and lipids. Augmented levels of dopamine in the synapse induced by METH have been shown in humans (Wilson et al., 1996; McCann et al., 1998), non-human primates (Seiden et al., 1976; Villemagne et al., 1998; Harvey et al., 2000) and rodents (Wagner et al., 1980; Yu et al., 2002). These studies also show that METH causes toxicity of the striatal nerve terminals. Administration of amphetamines and its analogs like METH results in long-lasting neurochemical depletions in dopamine and serotonin levels, inhibition of tyrosine hydroxylase and tryptophan hydroxylase activity, and dopamine reuptake sites (Davidson et al., 2001; Hanson et al., 2004; McCann and Ricaurte, 2004). In addition to the damage at the dopamine terminals, there is emerging evidence demonstrating that METH causes injury to cell bodies in various brain regions. For example, significant cortical gray matter and hippocampal deficits were seen in the brains of humans who used METH (Thompson et al., 2004). Rodents treated with METH displayed Fluoro-Jade B-positive degenerating neurons in the striatum (Yu et al., 2004). Repeated exposure to METH resulted in cell body injury in the parietal cortex of rats (Eisch and Marshall, 1998). In addition, a number of studies demonstrated the induction by METH of terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling (TUNEL)-positive apoptotic cells in the cortex and striatum of rodents (Stumm et al., 1999; Deng and Cadet, 2000; Deng et al., 2001; Zhu et al., 2002; Cadet et al., 2003; Loonam et al., 2003). In view of the neurotoxic effects of METH and the selective vulnerability of some striatal neurons toward dopaminergic manipulation, it is important to determine the phenotype of striatal neuron killed by an acute administration of METH.

The striatum is the largest nucleus of the basal ganglia and is associated with disorders of movement (Hickey and Chesselet, 2003; Singer and Minzer, 2003). There is increasing evidence that it is also involved in the control of attention, executive function, motivated behaviors (Alexander et al., 1986) as well as neuropsychiatric conditions such as obsessive compulsive disorders, psychoses and addictive behaviors (Calabresi et al., 1997). This structure is heterogeneous in its internal organization as well as in its afferent and efferent connections. Approximately 90% of the neurons of the striatum are GABAergic medium spiny projection neurons that are further subdivided into striatonigral projections containing substance P and dynorphin and striatopallidal projections containing enkephalin. The remaining 10% make up various striatal interneurons, of which the GABA-parvalbumin, somatostatin (SST), and cholinergic interneurons are considered the most prevalent (Gerfen, 1992). The selective loss of some striatal neurons has been associated with some neurodegenerative disorders. For example, striatal atrophy and choreiform movements seen in Huntington's disease are attributed to the progressive degeneration of striatal medium spiny projection neurons (Reiner et al., 1988; Albin et al., 1990). Considerable evidence now shows that it is the enkephalin-positive striatopallidal neurons that are the most vulnerable (Albin et al., 1992; Richfield et al., 1995, 2002; Mitchell et al., 1999; Deng et al., 2004) while striatal aspiny interneurons are spared in this disease (Ferrante et al., 1985; Reiner et al., 1988; Cicchetti et al., 1996; Hickey and Chesselet, 2003). Excitotoxic lesions by quinolinic acid are associated with increased concentrations of SST and neuropeptide Y as a result of preferential sparing of striatal interneurons (Beal et al., 1986, 1991). Damage to the striatum by the mitochondrial inhibitor, 3-nitopropionic acid, is specific to projection neurons while dopaminergic axons and nicotinamide adenine dinucleotide phosphate (NADPH)-diaphorase positive interneurons are spared (Guyot et al., 1997). Interneurons that contain NADPH-diaphorase, SST and neuropeptide Y are also less vulnerable in models of ischemia (Uemura et al., 1990). In the light of these findings, excessive METH use poses a serious health concern because the neuronal deficits in the aftermath of METH may lead to the eventual impairment of normal striatal function.

In the present study we employed immunohistofluorescence to assess the impact of METH on selective neuronal populations of the striatum. We combined TUNEL staining with immunohistofluorescence for SST, choline acetyltransferase (ChAT), and parvalbumin to assess the interneurons, and dopamine- and cAMP-regulated phosphoprotein, of apparent M_r 32,000 (DARPP-32) for the projection neurons. Our results demonstrate that the projection neurons, the cholinergic, and the GABA-parvalbumin neurons are decreased in number in the striatum after METH.

EXPERIMENTAL PROCEDURES

Animals and drug administration

A single i.p. injection of METH (10, 20, 30 or 40 mg/kg of body weight; Sigma, St. Louis, MO, USA) was given to male ICR mice 10–11 weeks of age (Taconic, Germantown, NY). All animals were housed individually with food and water available ad libitum on a 12-h light/dark cycle. Animals were habituated for two weeks prior to any drug treatment. Twenty-four hours post-treatment, animals are either killed by decapitation or first anesthetized with ketamine/acepromazine (100 mg/kg, 3 mg/kg of body weight) and then perfused transcardially with phosphate-buffered saline, pH 7.5 (PBS), containing 5000 U/ml of heparin followed by 4% paraformaldehyde in PBS. The brains were then either dissected out and immediately placed on dry ice or post-fixed in 4% paraformaldehyde in PBS overnight and cryoprotected in 30% sucrose in PBS at 4 °C. The tissue was then frozen and stored at −80 °C until use. Procedures of animal use were according to the National Institutes of Health Guide for the Care and Use of Laboratory Animals and were approved by the Institutional Animal Care and Use Committee of Hunter College of the City University of New York. All efforts were made to minimize the number of animals used and their suffering.

TUNEL histochemistry

The method was as described by Xu et al. (2005) with minor modifications. In brief, fresh frozen 20 μm coronal sections were collected between bregma 0.38 mm±0.1 mm and fixed in 4% paraformaldehyde for 30 min. After a wash with PBS, pH 7.6, the sections were immersed in 0.4% Triton-X-100 in PBS for 5–10 min at 70 °C. Sections were washed and TUNEL reactions (Roche Applied Science, Indianapolis, IN, USA) were applied directly onto sections and incubated for 1 h in a humidified chamber. After TUNEL staining, sections were counterstained with DAPI. Stained sections were washed in PBS and overlaid with a glass coverslip with Vectashield (Vector Laboratories, Burlingame, CA, USA). Images were taken using Molecular Dynamics CLSM Multiprobe 2001 scanning confocal system (Molecular Dynamics, Sunnyvale, CA, USA) or with a Nikon Eclipse E400 epifluorescent scope (Nikon, Melville, NY, USA) attached to a Hamamatsu digital camera C4742-95 (Hamamatsu, Bridgewater, NJ, USA) using FITC filters.

Double-labeling immunohistofluorescence

Free-floating serial coronal sections 20 μm in thickness were collected between bregma 0.38 mm±0.1 mm. The tissue was washed in PBS and then immersed in heated 0.4% Triton X-100 in PBS for 5 min and immediately washed in PBS. For neuron-specific nuclear protein (NeuN) or parvalbumin immunohistofluorescence, nonspecific binding sites were blocked with M.O.M.^® Mouse Ig Blocking Reagent (Vector Laboratories) for 1 h. Sections were washed and incubated with M.O.M.^® diluent for 15 min. Sections were then incubated in mouse anti-NeuN (1:500, Chemicon, Temecula, CA, USA) or mouse anti-parvalbumin (1:2000, Chemicon) in M.O.M^® diluent in 0.1% Triton-X-100 in PBS overnight at 4 °C. Sections were washed in PBS and blocked again with 5% normal goat serum in 0.1% Triton X-100 in PBS for 30 min. After PBS washes, sections were incubated with Cy3-conjugated goat anti-mouse (1:1000, Chemicon) in 1% normal goat serum/0.1% Triton X-100 in PBS for 1 h and washed with PBS.

For SST or DARPP-32 immunohistofluorescence, nonspecific binding sites were blocked with 10% normal goat serum and 0.2% Triton X-100 in PBS for 1 h. Sections were washed in PBS and incubated with rabbit anti-SST (1:500, Chemicon) in 2% normal goat serum/0.1% Triton X-100 in PBS or rabbit anti-DARPP-32 (1:200, Cell Signal) overnight at 4 °C. After a wash with PBS, sections were incubated with Cy3-conjugated goat anti-rabbit (1: 1000, Chemicon) in 1% normal goat serum/0.2% Triton X-100 in PBS for 1 h and rinsed with PBS.

For ChAT immunohistofluorescence, nonspecific binding sites were blocked with 10% normal rabbit serum/0.2% Triton X-100 in PBS for 1 h. Sections were washed in PBS and incubated with goat anti-ChAT (1:500, Chemicon) in 2% normal rabbit serum/0.2% Triton X-100 overnight. After a wash with PBS, the sections were incubated with Cy3-conjugated rabbit anti-goat (1: 1000, Chemicon) in 1% normal rabbit serum/0.2% Triton X-100 in PBS for 1 h and rinsed with PBS.

No staining was observed when primary antibody was left out or when primary antibody was pre-absorbed with either SST or ChAT. After a wash with PBS, the sections were incubated in the TUNEL reaction cocktail at 37 °C for 1 h. The sections were washed in PBS, mounted on slides, and overlaid with a coverslip with Vectashield (Vector Laboratories). All incubations and washes were performed at room temperature unless otherwise stated. Images were viewed and digitized with a Nikon Eclipse E400 epifluorescence microscope attached to a Hamamatsu digital camera C4742-95 using rhodamine and FITC filters.

Cell counts and quantification

Histological quantification of METH-induced apoptosis in striatal neurons was as described by Zhu et al. (2005). Briefly, cells of interest were quantified from 20 μm thick coronal sections in an area of 0.26 mm² for each aspect of the striatum (dorsal–medial [DM], dorsal–lateral [DL], ventral–medial [VM], and ventral–lateral [VL]) (see schematic diagram in Fig. 1 [adapted from Hof et al., 2000]). Average neuronal cell counts (neurons labeled with NeuN) were obtained from six control animals as a baseline for quantification of the percentage of TUNEL-positive neurons in experimental animals. TUNEL cell counts were averaged from five 20 μm serial sections per animal (see Fig. 2). The observer who did the cell counts and immunofluorescence quantitation was blind to treatment group.

Fig. 1.

Schematic of the striatum indicating the four regions selected for cell counts. DM, VM, DL, and VL. Black boxes indicate the 0.26 mm² area where cells were manually counted after immunostaining. Reproduced from Hof et al. (2000).

Fig. 2.

Estimation of the number of NeuN-positive neurons in the four quadrants of the striatum. Coronal sections through the striatum (bregma 0.38±0.1 mm) were processed for immunohistofluorescence with Cy3-labeled antibodies against the NeuN. NeuN-positive neurons were counted from 20 μm thick coronal sections in an area of 0.26 mm² for each of the four quadrants of the caudate-putamen (CPu, DM, DL, VM, and VL). The mean number of NeuN-positive neurons was taken from an average of 30 sections from six control animals (five serials sections from each animal). No significant deviation in total cell counts was observed between animals or between regions of the striatum.

Nissl staining

Coronals sections of the striatum from bregma 0.38 mm±0.1 mm were fixed and defatted. Slides were then rinsed with water followed by 2 min incubation in 0.2% Cresyl Violet. Sections were then washed with water and dehydrated in a graded series of alcohol/water solutions. Tissue was placed in xylene for 5 min and a coverslip was gently overlaid.

Statistical analysis

Analysis is performed from mean±S.E.M. Differences between groups were analyzed by ANOVA followed by post hoc comparison using Fisher's protected least significance test. Significance criteria were set at P<0.05. Student's t-test was used to compare differences between two groups.

RESULTS

METH induces apoptosis in some neurons of the striatum

We subdivided the striatum into four compartments in order to assess potential differences between dorsal–ventral and lateral–medial aspects of this structure (Fig. 1). To determine the percentage of neurons undergoing METH-induced apoptosis, we first established a consistent and reliable baseline of the total number of neurons within each aspect of the striatum. This was accomplished by labeling the neurons with an antibody against the neuron-specific marker NeuN, followed by a secondary antibody conjugated to the chromophore Cy3. Five striatal sections (bregma 0.38 mm±0.1 mm) from each animal (total of six animals) were processed for immunohistofluorescence for NeuN and all fluorescent neurons were counted manually under the microscope. The average number of NeuN-positive neurons from the 30 sections was taken from each region of interest (Fig. 2) and later used to determine the percentage of TUNEL-positive neurons in the aftermath of METH. The total number of NeuN immunopositive neurons in each aspect of the striatum was very consistent between animals (Fig. 2).

We tested various doses of METH from 10 to 40 mg/kg of body weight on the induction of TUNEL-positive nuclei in the striatum. We chose TUNEL staining (a late marker of apoptosis) because it has been shown that some apoptosis can occur independently of activation of caspase-3, an early marker of apoptosis (Ferrer et al., 2003; Spalding et al., 2005). There was a sharp increase in the amount of TUNEL-positive nuclei going from 20 to 30 mg/kg, i.p. (Fig. 3A). Analysis of cell death by METH over time showed that TUNEL positive cells peaked 24 h post-treatment (Zhu et al., 2005). Continued increase of dosing at 40 mg/kg did not show any significant increase from 30 mg/kg. Thus, a 30 mg/kg dose of METH was chosen for subsequent studies. Although the DM aspect of the striatum consistently displayed lower levels of METH-induced apoptosis relative to other striatal quadrants, no significance was found between the four different regions of the striatum (Fig. 3B). The average amount of apoptosis induced by METH (30 mg/kg) was approximately 25%; however, there was a large inter-animal variability in the amount of METH-induced TUNEL-positive cells (Fig. 3C). We measured body core temperature and found that METH-induced hyperthermia does not display a large inter-animal variability (Zhu et al., 2006).

Fig. 3.

Induction of cell death by METH as a function of dose. Mice were injected once with METH (i.p.) at the doses indicated. The mice were killed 24 h after METH and sections of striatal tissue were processed for TUNEL immunohistofluorescence. (A) Scanning confocal micrographs of TUNEL staining in the striatum of mice treated with increasing doses of METH (a–e). Scale bar=50 μm. (B) Percentage of striatal neurons displaying TUNEL staining for DM, VM, DL, and VL regions of the striatum (mean S.E.M.). * P<0.005, ** P<0.01, *** P<0.05 compared with saline control. (C) Scatter graphs indicate percentage of TUNEL-positive neurons for each animal within each treatment group. Each dot represents one animal. Blue=saline, red=METH 10 mg/kg, green=METH 20 mg/kg, purple=METH 30 mg/kg, yellow=METH 40 mg/kg; n=10–11 per experimental group.

To determine whether METH induced apoptosis occurs selectively in neurons of the striatum instead of glial cells, we combined TUNEL with immunohistofluorescence for NeuN. This marker was chosen as opposed to other neuronal markers because the labeling of this marker occurs primarily in the nucleus and TUNEL-labeling is also a nuclear stain. Immunohistofluorescence showed that TUNEL-positive nuclei colocalized with NeuN (Fig. 4g–i). METH treatment caused NeuN staining to be less pronounced (Fig. 4d), probably an indication that these neurons have already incurred severe deficits in molecular markers such as receptors and peptides. We did not assess TUNEL with glial-specific stains because all the TUNEL staining colocalized with immunohistofluorescence for NeuN.

Fig. 4.

METH induces apoptosis in some striatal neurons of the striatum. Double-labeled epifluorescent micrographs of striatal tissue stained with Cy3-labeled antibodies against NeuN and TUNEL with FITC-conjugated dUTPs in control (a–c) and METH-treated (30 mg/kg, i.p., killed 24 h post-treatment) animals (d–i). Bottom panels (c, f, i) are overlays of both TUNEL and NeuN staining. Higher magnification of METH-treated animals indicates that NeuN-positive neurons overlap with TUNEL-positive cells (g–i). White arrows point to overlapping TUNEL and NeuN positive cells. Scale bar=100 μm (a–f), 20 μm (g–i).

Effects of METH on the different neuronal populations of the striatum

GABAergic projection neurons

We tried to combine TUNEL with immunohistofluorescence for various selective markers of the striatonigral and striatopallidal projection neurons; however, the results were unsatisfactory. For example, we attempted to colocalize TUNEL with immunohistofluorescence for dopamine D1 or D2 receptors and the neuropeptides substance P, dynorphin, and enkephalin (data not shown). Additionally, we injected a fluorescent retrograde tracer into the terminal field areas of striatopallidal (globus pallidum) and striatonigral (substantia nigra pars reticulata) projections but this approach was also unsuccessful (data not shown). We believe that all these attempts were unsuccessful because by the time some striatal neurons displayed METH-induced TUNEL staining (extensive nicking of the nuclear DNA), the levels of molecular markers such as receptors and neuropeptides reached levels below the limit of detection of immunohistochemistry with the commercial antibodies that we employed. In the light of these technical constraints, we decided to demonstrate the apoptosis of projection neurons by an indirect method using immunohistofluorescence for DARPP-32, a reliable marker for both striatopallidal and striatonigral projection neurons.

We did not observe colocalization of DARPP-32 with TUNEL-positive cells (Fig. 5A g–i). However, immunoreactivity of DARPP-32 showed significant decreases in the METH-treated group compared with control (Fig. 5A a–f). Closer observation showed that in a given area where there was TUNEL staining, it was very difficult to also find DARPP-32 staining (Fig. 5A d–f). This was dramatically different in controls where there was abundant DARPP-32 staining and no TUNEL (Fig. 5A a–c). Cell counts of DARPP-32-positive neurons showed that METH caused a 15.4%±2.5% decrease of DARPP-32-positive neurons in the DM region, 23.8%±3.8% in the VM region, 22.8%±3.8% in the VL region and 24.4%±3.9% in the DL region compared with controls (Fig. 5B). This decrease corresponded with the increases in percentage of TUNEL-positive neurons in Fig. 3B. The immunohistological approach used here demonstrated that the loss of DARPP-32 correlated with the increases in TUNEL-positive cells, consistent with the interpretation that METH induces the loss of some projection neurons of the striatum.

Fig. 5.

The number of DARPP-32-containing projection neurons is decreased by METH. Mice received one injection of METH (i.p.) at 30 mg/kg of body weight and were killed 24 h later. Sections of striatal tissue were processed for immunohistofluorescence. (A) Double-labeled epifluorescent micrographs of striatal tissue stained with Cy3-labeled antibodies against DARPP-32 and TUNEL with FITC-conjugated dUTPs in control (a–c) and METH-treated animals (d–i). Bottom panels (c, f, I, l) are overlays of both TUNEL and DARPP-32 staining. Higher magnification of METH-treated animals indicates that the two chromophores do not overlap (g–i). Scale bar=100 μm (a–f), 20 μm (g–i). (B) Counts of DARPP-32-positive neurons (mean±S.E.M.) demonstrate a significant decrease in all four quadrants of the striatum after exposure to METH. * P<0.05 compared with corresponding regions of saline control. CPu, caudate-putamen.

Parvalbumin-containing GABA interneurons

We did not observe colocalization of immunohistofluorescence of parvalbumin-containing GABAergic interneurons and TUNEL-positive neurons (Fig. 6A j–l). Despite this, immunoreactivity of parvalbumin-positive neurons was diminished (Fig. 6A d–f) and dendritic arbors appeared shorter (Fig. 6A j–l) in METH-treated animals compared with controls (Fig. 6A a–c, g–i). To detect the potential loss of some of these interneurons after exposure to METH, cell counts of parvalbumin-positive neurons were performed in the four quadrants of the striatum. Although all four quadrants showed a decrease, it was the dorsal region of the striatum where this GABAergic interneuron was concentrated and most affected by METH treatment; causing a 43.7%±5.5% decrease of the marker in the DM region and 49%±10.8% in the DL region (Fig. 6B).

Fig. 6.

Immunohistofluorescence of parvalbumin interneurons is diminished by METH. Mice received a single injection of METH (i.p.) at a dose of 30 mg/kg and were killed 24 h after the treatment. Sections of striatal tissue were processed for immunohistofluorescence. (A) Double-labeled epifluorescent micrographs of striatal tissue stained with Cy3-labeled antibodies against parvalbumin and TUNEL with FITC-conjugated dUTPs in control (a–c) and METH-treated animals (d–f). Bottom panels (c, f, i, l) are overlays of both TUNEL and parvalbumin staining. Higher magnification of tissue from METH-treated animals demonstrated that the two chromophores do not overlap (g–i), although cell bodies showed significant loss of dendritic arborizations (j–l). Scale bar=100 μm (a–f), 20 μm (g–l). (B) Counts of parvalbumin-positive neurons (mean±S.E.M.) demonstrate a significant decrease of immunohistofluorescence in the dorsal regions of the striatum. * P<0.0001, ** P<0.005, *** P<0.05 compared with corresponding regions of saline control. ^! P<0.0001, ^!! P<0.001 compared with the DM region of saline. ^† P<0.005 compared with the DL region of saline. CPu, caudate-putamen.

Cholinergic and SST interneurons

Colocalization of immunohistofluorescence between ChAT-containing cholinergic interneurons and TUNEL-positive neurons was not observed throughout the striatum. No obvious change was seen from the histology in the number and morphology of the cholinergic interneurons after METH treatment compared with control. Cell counts of the cholinergic interneurons confirmed these neurons to be unaffected in the ventral portions of the striatum. However, the dorsal regions of the striatum showed loss of this cell type. This is most significant for the DM portion of the striatum, which had a 28.9%±4.8% decrease from control (Fig. 7B). It should be noted that these deficits were observed in the animals that displayed high amounts of TUNEL.

Fig. 7.

Some cholinergic interneurons are vulnerable to METH. Animals were injected once with METH (i.p., 30 mg/kg) and killed 24 h after METH. Coronal sections through the striatum were processed for immunohistofluorescence. (A) Double-labeled epifluorescent micrographs of striatal tissue stained with Cy3-labeled antibodies against ChAT and TUNEL with FITC-conjugated dUTPs in control (a–c) and METH-treated mice (d–i). Bottom panels (c, f, i) are overlays of both TUNEL and ChAT staining. Higher magnification of METH-treated group demonstrates the lack of colocalization between the two chromophores (g–i). Scale bar=100 μm (a–f), 20 μm (g–i). (B) Counts of ChAT-positive neurons (mean±S.E.M.) demonstrate a significant decrease in the DM region of the striatum. * P<0.005 compared with the corresponding regions of saline control. ^! P<0.0001, ^† P<0.005 compared with the DL region of saline. ^!! P<0.0005 compared with DM region of saline. ^# P<0.05 compared with the DM of METH. ^% P<0.05 compared with the DL region of METH. CPu, caudate-putamen.

No colocalization of immunohistofluorescence was found between SST and TUNEL-positive neurons (Fig. 8A). Like the cholinergic interneurons, there is no obvious morphological change observable throughout the striatum. Cell counts in the four quadrants of the striatum did not reveal the loss of immunohistofluorescence for SST (Fig. 8B), suggesting that this population of striatal interneuron is refractory to METH.

Fig. 8.

SST interneurons are refractory to METH-induced apoptosis. (A) Double-labeled epifluorescent micrographs of striatal tissue stained with antibody against SST (Cy3) and TUNEL with FITC-conjugated dUTPs in control (a–c) and METH-treated (30 mg/kg, i.p., killed 24 h post-treatment) mice (d–i). Bottom panels (c, f, i) are overlays of both TUNEL and SST staining. Higher magnification of METH-treated mice shows absence of overlap between the two chromophores (g–i). Scale bar=100 μm (a–f), 20 μm (g–i). (B) Counts of SST-positive neurons (mean±S.E.M.) demonstrate no statistical significance between the two treatment groups or between the regions within the treatment groups. CPu, caudate putamen.

Effect of METH on Nissl staining of striatal tissue

Because we did not observe colocalization between TUNEL and selective markers of striatal neurons, we stained striatal sections from control and METH-treated animals with Cresyl Violet. We found a significant decrease in the number of cells stained with Cresyl Violet in the METH-treated group (Fig. 9A and B). The METH-treated animals showed cells that appeared to have fragmented and clumped nuclei (Fig. 9A). The magnitude of cell loss observed with the Nissl stain correlated with the amount of TUNEL staining (compare Figs. 3B and 9B). Moreover, in all sections examined, a high density of TUNEL always appeared with a paucity of DARPP-32 staining. These observations demonstrate the loss of some striatal neurons in the aftermath of METH.

Fig. 9.

Effect of METH treatment on the cell density of the striatum. Mice were treated with a single METH 30 mg/kg (i.p.) injection and killed 24 h later. Coronal sections of the striatum were stained with Cresyl Violet. Light micrographs of Nissl staining in saline (A) and METH 30 mg/kg (B) treated animals. (C) Nissl-stained cells were counted in 20 μm thick coronal sections of the striatum in an area of 0.26 mm² for each of the four quadrants of the caudate-putamen (CPu, DM, DL, VM, and VL). Results represent the mean±S.E.M. number of cells counted in 10 animals per experimental group. * P<0.05 compared with controls of each corresponding striatal region. No significance was found between regions.

DISCUSSION

None of the selected striatal markers for medium spiny neurons (DARPP-32) and interneurons (ChAT, SST and parvalbumin) exhibited colocalization with the TUNEL-positive cells. This would initially suggest that these neuronal cells were not undergoing apoptosis. But the reduction in population size for these cells is observed only in animals showing TUNEL-positive cells, suggesting that some striatal neurons die by apoptosis after METH. Furthermore, the magnitude of cell loss observed with Nissl-stained sections correlated with the amount of cell loss observed with TUNEL staining. Labeling by TUNEL marks the occurrence of DNA fragmentation, which occurs late in the apoptosis process. Thus, damage to neuronal markers may have occurred before DNA fragmentation. The agreement between TUNEL and the loss of Nissl staining in striatal tissue of animals treated with METH demonstrates that some neurons are lost in the aftermath of METH.

METH is a potent psychostimulant that induces excessively high levels of dopamine in the extracellular compartment. Excess release of endogenous striatal dopamine has been hypothesized to mediate neurotoxicity at dopaminergic terminals and cause neuronal damage to some striatal neurons (Davidson et al., 2001). Metabolites of dopamine generate toxic free radicals, which contribute to this damage (Cadet and Brannock, 1998). The connection between METH-induced dopamine terminal toxicity and elevated dopamine levels was established using pharmacological agents that blocked post-synaptic dopamine receptors or pre-synaptic dopamine transporters, such agents were found to protect from METH (Buening and Gibb, 1974; Gibb and Kogan, 1979; Hotchkiss and Gibb, 1980; Sonsalla et al., 1986; O'Dell et al., 1993; Xu et al., 2005). Microinjections of the dopamine agonist, SKF38393, administered centrally into the rat brain are toxic to striatal neurons (Kelley et al., 1990). Additionally, we have shown that blockade of dopamine receptors using the D1 receptor antagonist, SCH23390, or D2 receptor antagonist, raclopride, prior to METH administration can prevent apoptosis (Xu et al., 2005). Since excess dopamine is neurotoxic, the local anatomical organization of dopamine terminals and receptors plays an important role in where damage occurs following METH administration. Given that relatively high levels of dopamine receptors are located on the medium spiny neurons, we can thus speculate that these neurons would be more vulnerable to the deleterious effects of METH administration. Another type of neurotransmission that could account for the differential vulnerability of striatal neurons to METH involves the neurotransmitter glutamate.

Dopaminergic and glutamatergic systems display complex anatomical arrangements in the striatum. Microdialysis studies show that METH induces a delayed release of glutamate in the striatum (Nash and Yamamoto, 1992). Agents that block this delayed increase of extracellular glutamate can prevent METH-induced depletion of dopamine content (Stephans and Yamamoto, 1994). Glutamate analogues such as quinolinic acid and kainic acid have been shown to induce excitotoxic lesions of striatal neurons (Beal et al., 1991; Qin et al., 1996). Excitotoxicity is mediated by ionotrophic NMDA and AMPA/KA receptors (Beal, 1992; Nakanishi, 1992). Excessive activation of glutamate receptors is thought to contribute to neuronal death in neurological disorders such as trauma, epilepsy, Parkinson's disease, and Alzheimer's disease (Beal, 1992; Marino et al., 2003; Arundine and Tymianski, 2004; Hynd et al., 2004). In addition, activation of the NMDA receptor in the striatum leads to the activation of nitric oxide synthase (Szabo, 1996), which results in the synthesis of the diffusible second messenger nitric oxide, an agent that has been associated with damage of surrounding neurons in the brain (Dawson et al., 1991; Dawson and Snyder, 1994). Given the reciprocal interactions between dopamine and glutamate, the selective vulnerability of some striatal neuron subtypes may be due to the anatomical differences in the expression of glutamate receptor subtypes and subunits. Studies indicate that the majority of glutamate receptors are found on medium spiny neurons. These neurons contain the NMDA NR1 and NR2A/2b subunits in the striatum as well as AMPA GluR1, GluR2/3, and GluR4 subunits (Chen and Reiner, 1996; Chen et al., 1998; Hu et al., 2004; Wang et al., 2004). Cholinergic interneurons and half of the GABAergic parvalbumin-containing interneurons express NMDA NR2A/2B R while SST interneurons do not (Landwehrmeyer et al., 1995; Chen and Reiner, 1996; Chen et al., 1996). The different responses to glutamate receptor activation are reflected in the localization of these receptors within the striatum. Given that SST interneurons lack the receptor, it may account for the resistance to NMDA-mediated excitotoxicity on these neurons. Furthermore, damage to medium spiny neurons and parvalbumin interneurons may be due to the presence of these receptors.

The various populations of striatal neurons are intricately interconnected affecting each other within the striatum and at the targets of innervation of the medium spiny projection neurons. Thus, it is reasonable to assume that the loss of some projection and interneurons after METH may have an impact on the surviving neighboring neurons. The projection neurons exert inhibition on their targets in the substantia nigra and the globus pallidum via GABA_A receptors (Precht and Yoshida, 1971). In addition, the GABAergic medium spiny neurons can inhibit each other via their recurrent collaterals within the striatum (Wilson and Groves, 1980; Yung et al., 1996). In the light of this, the loss of some projection neurons after METH could be interpreted to result in attenuation of inhibition within the striatum and in the targets of the projection neurons. However, the state of the projection neurons of the striatum is affected by the activity of the interneurons. Our results demonstrate that some cholinergic and GABA-parvalbumin interneurons may also be killed by acute METH.

Our data suggest that approximately half of the GABA-parvalbumin interneurons of the dorsal striatum are killed by METH. These interneurons have a high content of GABA (Bolam et al., 1983) and receive glutamatergic excitation from the cortex (Koos and Tepper, 1999). Thus, in the aftermath of cortical transmission these interneurons exert inhibition on striatal neurons via GABA, and it is noteworthy that METH ablates nearly half of this population of interneuron in the dorsal striatum. We demonstrate here that another population of striatal interneuron impacted by METH is the cholinergic interneuron. These interneurons affect the output of the projection neurons. For example, cholinergic transmission augments striatal dopamine release (Zhou et al., 2001) and increases the responses of the medium spiny projection neurons to glutamate (Calabresi et al., 1998). The present study demonstrates that in the post-METH striatum the interneuron balance is skewed in favor of transmission by the SST/nitric oxide synthase interneurons because the latter population is refractory to METH. The nitric oxide made by the SST/nitric oxide synthase interneurons is known to increase the release of glutamate and dopamine in the striatum (Kawaguchi et al., 1995). The striatal SST interneurons also express the neuropeptide NPY (Kawaguchi et al., 1995). These neuropeptides may play a neuroprotective role in the striatum. For example, SST significantly attenuated ischemic neuronal damage induced by middle cerebral artery occlusion in the rat brain (Rauca et al., 1999). Similarly, SST attenuated NMDA-induced cell death in cultured fetal rat cortical neurons (Forloni et al., 1997). In addition, a recent study showed that the intraventricular administration of the neuropeptide NPY protected the mouse striatum from METH-induced apoptosis (Thiriet et al., 2005). These studies suggest that SST and NPY belong to a class of neuroprotective peptides in the brain. It needs to be investigated how these neuropeptides confer protection on the neuron expressing them, since our results demonstrate that this type of interneuron is resistant to METH-induced damage.

CONCLUSION

In summary, our results demonstrate the loss of some striatal neurons in the aftermath of METH. This conclusion is supported by the appearance of TUNEL staining in the striatum, the loss of immunostaining for selective markers of striatal neurons, and decreased density of Nissl staining after METH. The affected phenotypes are: projection neurons, cholinergic and GABA-parvalbumin interneurons. Of note is the observation that the SST/NPY interneurons are refractory to METH. More work is needed to investigate the functional state, namely the output, of the striatal neurons that survive the ravages of METH.

Abbreviations

ChAT

choline acetyltransferase

DARPP-32

dopamine- and cAMP-regulated phosphoprotein, of apparent M_r 32,000

DL

dorsal–lateral

DM

dorsal–medial

METH

methamphetamine

NADPH

nicotinamide adenine dinucleotide phosphate

NeuN

neuron-specific nuclear protein

PBS

phosphate-buffered saline

SST

somatostatin

TUNEL

terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling

VL

ventral–lateral

VM

ventral–medial