L-type Ca2+ channels (LTCCs) play an important role in chronic psychostimulant-induced behaviors. However, the Ca2+ second messenger pathways activated by LTCCs after acute and recurrent psychostimulant administration that contribute to drug-induced molecular adaptations are poorly understood. Using a chronic amphetamine treatment paradigm in rats, we have examined the role of LTCCs in activating the mitogen-activated protein (MAP) kinase pathway in the ventral tegmental area (VTA), a primary target for the reinforcing properties of psychostimulants. Using immunoblot and immunohistochemical analyses, we find that in chronic saline-treated rats a challenge injection of amphetamine increases phosphorylation of MAP [extracellular signal-regulated kinase 1/2 (ERK1/2)] kinase in the VTA that is independent of LTCCs. However, in chronic amphetamine-treated rats there is no increase in amphetamine-mediated ERK1/2 phosphorylation unless LTCCs are blocked, in which case there is robust phosphorylation in VTA dopamine neurons. Examination of the expression of phosphatases reveals an increase in calcineurin [protein phosphatase 2B (PP2B)] and MAP kinase phosphatase-1 (MKP-1) in the VTA. Using in situ hybridization histochemistry and immunoblot analyses, we further examined the mRNA and protein expression of the LTCC subtypes Cav1.2 and Cav1.3 in VTA dopamine neurons in drug-naive animals and in rats after chronic amphetamine treatment. We found an increase in Cav1.2 mRNA and protein levels, with no change in Cav1.3. Together, our results suggest that one aspect of LTCC-induced changes in second messenger pathways after chronic amphetamine exposure involves activation of the MAP kinase phosphatase pathway by upregulation of Cav1.2 in VTA dopaminergic neurons.
Repeated exposure to the psychostimulants amphetamine and cocaine causes long-lasting neuronal adaptations that underlie compulsive behavior (Hyman and Malenka, 2001; Nestler, 2001b; Robinson and Berridge, 2003). The ventral tegmental area (VTA) serves as a primary neural site that initiates the molecular mechanisms that underlie such behaviors (Vezina and Stewart, 1990; Vezina et al., 2002; Bonci et al., 2003). Amphetamine and cocaine activate intracellular signaling pathways in the VTA that cause downstream changes in protein phosphorylation and gene expression (White and Kalivas, 1998; Wolf, 1998; Licata and Pierce, 2003). In addition to the pivotal role of Ca2+ signaling via the NMDA subtype of the glutamate receptor, Ca2+ influx via voltage-gated L-type Ca2+ channels (LTCCs) is necessary for psychostimulant-induced behavioral and neurochemical changes (Karler et al., 1991; Pani et al., 1991; Kuzmin et al., 1992; Suzuki et al., 1992; Martellotta et al., 1994; Pierce and Kalivas, 1997; Pierce et al., 1998; Licata et al., 2000, 2001; Pliakas and Carlezon, 2001). However, their respective contribution to psychostimulant-induced changes remains unknown. Based on pharmacological studies, LTCCs appear to play a downstream role in NMDA and psychostimulant-induced sensitization. NMDA blockers inhibit both acute and chronic psychostimulant-induced behaviors (Karler et al., 1989; Vezina and Queen, 2000), whereas LTCC blockers inhibit only chronic, but not acute, psychostimulant-induced neurochemical and behavioral changes (Karler et al., 1991; Pierce and Kalivas, 1997; Pierce et al., 1998).
LTCCs play an important role in long-term neuronal plasticity (Bito et al., 1996; Deisseroth et al., 2003). They activate both Ca2+-mediated kinase and phosphatase pathways (Deisseroth et al., 2003; Groth et al., 2003). LTCCs are heteromeric complexes and form channels that contain either the Cav1.2 or Cav1.3 Ca2+ pore-forming subunit (Catterall, 2000; Ertel et al., 2000; Stotz and Zamponi, 2001). Both Cav1.2 and Cav1.3 are activated and blocked by the same drugs and hence traditionally are believed to have similar properties. However, there is growing evidence that Cav1.2 and Cav1.3 have different neuronal distributions (Hell et al., 1993; Furuyashiki et al., 2002) and different physiological characteristics (Koschak et al., 2001; Xu and Lipscombe, 2001; Lipscombe, 2002). Cav1.2-containing channels open at high voltages, are present in signaling complexes (Davare et al., 1999, 2000, 2001), and activate Ca2+-mediated second messenger pathways such as the mitogen-activated protein (MAP) kinase pathway (Dolmetsch et al., 2001; Weick et al., 2003). In contrast, the Cav1.3-containing LTCCs activate at low voltages (Platzer et al., 2000; Koschak et al., 2001; Xu and Lipscombe, 2001) and are implicated in more rapid neuronal responses.
Here we present data for the role of LTCCs and their subtypes Cav1.2 and Cav1.3 in chronic amphetamine-mediated molecular changes in the VTA. We have used an amphetamine sensitization paradigm known to induce robust, long-lasting locomotor sensitization in rats (Wolf and Jeziorski, 1993; Wolf et al., 1994). We have examined the MAP kinase pathway, a mediator of psychostimulant-induced plasticity (Pierce et al., 1999; Valjent et al., 2000), and its regulation after acute and chronic amphetamine treatment.
Materials and Methods
Animals. Male Sprague Dawley rats (Charles River Laboratories, Wilmington, MA) weighing 200-225 gm at the start of the experiment were used in these studies. Rats were group housed and provided food and water ad libitum. A 12 hr light/dark cycle was used (from 7 A.M. to 7 P.M). All procedures were conducted in accordance with the Massachusetts General Hospital Subcommittee on Research Animal Care rules and regulations.
Drugs and antibodies. d-Amphetamine sulfate and diltiazem were obtained from Sigma (St. Louis, MO). Anti-mouse phospho-extracellular signal-regulated kinase 1/2 (ERK1/2 kinase), calcineurin-A, and tyrosine hydroxylase (TH) antibodies were obtained from Sigma and anti-rabbit ERK1/2 kinase antibody from Cell Signaling Technology (Beverly, MA). Anti-rabbit Cav1.2, Cav1.3, TH, and anti-mouse actin antibodies were from Chemicon (Temecula, CA). Anti-rabbit MAP kinase phosphatase 1 (MKP-1) and anti-goat MAP kinase phosphatase 3 (MKP-3) antibodies were from Santa Cruz Biotechnology (Santa Cruz, CA). Protease and phosphatase inhibitor mixtures were from Sigma.
Amphetamine treatment paradigm. The amphetamine treatment regimen chosen has been shown to induce robust induction of locomotor sensitization (Wolf and Jeziorski, 1993; Wolf et al., 1994). Rats were handled for 3-4 d. Rats were given intraperitoneal injections of saline or 5 mg/kg d-amphetamine sulfate once per day for 5 d. For the ERK1/2 kinase phosphorylation (P-ERK1/2) studies, on day 6 the rats were divided into acute and chronic treatment groups as outlined in Table 2. The acute group that was treated with saline on days 1-5 was challenged with saline or amphetamine with or without diltiazem pretreatment on day 6. The chronic amphetamine group was treated chronically with amphetamine on days 1-5 and challenged with saline or amphetamine with or without diltiazem pretreatment on day 6. For immunoblot analysis the rats were decapitated 15 min after the last injection, and for immunohistochemical analysis the rats were perfused 15 min after the last injection. For immunoblot analysis of all other proteins the rats were decapitated on day 6, 1 d after the last saline or amphetamine injection. For the in situ hybridization histochemistry the rats were decapitated 1, 3, or 14 d after the last injection.
Immunoblots. For the P-ERK1/2 immunoblots brains were dissected rapidly and frozen in isopentane at -40°C. Brains were sectioned in a cryostat to the rostral end of the nucleus accumbens (NAc), and a 1 mm section was cut with a razor blade. A 17 gauge steel bore was used to punch out both sides of the NAc. The same brain was sectioned to the rostral end of the VTA. A 1 mm section was cut, and an 18 gauge steel bore was used to get a midline punch of the VTA. Tissue samples were sonicated in SDS sample buffer [1% SDS in Tris-EDTA, pH 7.4, containing 1× protease inhibitor mixture [containing 1 mm AEBSF and (in μm) 0.08 aprotinin, 21 leupeptin, 36 bestatin, 15 pepstatin A, and 14 E-64], 5 mm sodium fluoride (NaF), 0.1 mg/ml benzamidine, and 1× phosphatase inhibitor mixture (containing cantharidin, bromotetramisole, microcystin-LR, sodium orthovanadate, sodium molybdate, sodium tartrate, and imidazole)]. Samples were boiled for 10 min and centrifuged at 14,000 rpm for 10 min. Supernatants were collected, and protein concentration was determined by the BCA assay (Pierce, Rockford, IL). Then 50 μg of protein was loaded on a 12% SDS-polyacrylamide gel. For Cav1.2 and Cav1.3 immunoblots the tissue was sonicated in TE, pH 7.4, containing the above protease and phosphatases inhibitors and also containing 8 μg/ml calpain inhibitor peptide; samples were separated on a 10% SDS-polyacrylamide gel. Protein was transferred to polyvinylidene fluoride (PVDF) membrane (Immuno-Blot PVDF, 0.2 μm, Bio-Rad, Hercules, CA) and blocked in blocking buffer (5% nonfat-dried milk in 0.25 m Tris-HCl, pH 7.6, 1.37 m NaCl, 0.1%Tween) for 60 min. Blots were probed with primary antibody (1:10,000, P-ERK1/2; 1:40,000, ERK1/2; and 1:1000, Cav1.2, Cav1.3, and actin) overnight at 4°C. The next day the blots were washed four times for 5 min each in blocking buffer. Blots were incubated with horse anti-mouse (1:15,000 for P-ERK1/2 and 1:10,000 for actin), goat anti-rabbit (1:20,000 for ERK1/2 and 1:10,000 for Cav1.2), or horse anti-goat (1:10,000 for Cav1.3) horseradish peroxidase-linked IgG (Vector Laboratories, Burlingame, CA) for 1 hr at room temperature. Blots were washed once for 15 min and four times for 5 min each. Protein bands were detected by chemiluminescence (Western Lightning, PerkinElmer Life Sciences, Boston, MA) and exposed to X-OMAT Blue autoradiographic film (Kodak, Rochester, NY). Kaleidoscope-prestained standards (Bio-Rad) were used for protein size determination. ERK1 and P-ERK1 were detected at 44 kDa; ERK2, P-ERK2, and MKP-3 at 42 kDa; MKP-1 at 38 kDa; protein phosphatase 2B (PP2B) at 61 kDa; and Cav1.2 and Cav1.3 at 210 kDa. For quantitation the blots were scanned with an HP ScanJet 7400c scanner (Hewlett Packard, Palo Alto, CA). Intensity of the protein bands was measured as optical density, using the NIH Image program (National Institutes of Health, Bethesda, MD).
Immunohistochemistry. At 15 min after the last injection the rats were perfused transcardially with 20 ml of 0.9% NaCl plus 0.1 mm NaF, followed by 300 ml of 4% paraformaldehyde (PFA) in 0.1 m phosphate buffer (PB; 0.1 m Na2HPO4/NaH2PO4, pH 7.5) with 0.1 mm NaF. Brains were removed and postfixed in 4% PFA for 2 hr and then placed in 15% sucrose in 0.1 m PB, pH 7.5, and 0.1 mm NaF at 4°C. Before sectioning, the brains were frozen in -40°C isopentane, and coronal sections (20 μm) were cut in a cryostat and kept in phosphate buffer, pH 7.5, with 0.1 mm NaF at 4°C. Sections were collected between -5.20 and -6.04 mm relative to bregma (Paxinos and Watson, 1997). Free-floating sections were rinsed in TBS (0.25 m Tris and 0.5 m NaCl, pH 7.5), incubated for 20 min at 4°C in methanol with 1% hydrogen peroxide, and then rinsed three times for 10 min each in TBS. Sections were incubated for 30 min in 3% horse serum and 0.2% Triton X-100 in TBS, followed by incubation in 1:500 phospho-ERK1/2 antibody for 24 hr at 4°C. The next day the sections were washed three times in TBS and incubated for 2 hr at room temperature with 1:200 biotinylated horse anti-mouse IgG (Vector Laboratories). Sections were incubated for 1 hr at room temperature in avidin-biotin-peroxidase complex solution (ABC solution, Vector Laboratories). Sections then were washed three times with TBS for 10 min each and developed for 12 min with 2 mg/ml 3,3′-diaminobenzidine tetrahydrochloride (DAB; Sigma). For quantitation, P-ERK1/2-positive neurons were counted bilaterally at a 20× magnification over a 0.25 mm2 area of the VTA.
For P-ERK1/2 and TH fluorescence double-labeling immunohistochemistry the sections were processed as above and incubated with 1:500 anti-mouse phospho-ERK1/2 and 1:1000 anti-rabbit TH antibody for 24 hr at 4°C. Sections were visualized with 1:200 anti-mouse Cy3 and anti-rabbit Cy2 secondary antibodies (Jackson ImmunoResearch, West Grove, PA) and mounted with Vectashield (Vector Laboratories). For the Cav1.2 and TH double-labeling experiments the sections were permeabilized with methanol and 1% H2O2 for 15 min. Sections were incubated at room temperature for 20 min with 5% donkey serum and 0.1% Triton X-100 in 0.1 m PB and then for 40 min with 5% bovine serum albumin in 0.1 m PB. Sections were incubated with 1:1000 anti-rabbit Cav1.2 and 1:1000 anti-mouse TH antibodies. For the P-ERK1/2 and Cav1.2 double labeling the sections were incubated with 1:500 anti-mouse P-ERK1/2 and 1:1000 anti-rabbit Cav1.2 antibodies. Sections were visualized with 1:200 anti-rabbit Cy3 and anti-mouse Cy2 secondary antibodies. Images were captured with the Zeiss Axiovert 200 fluorescence microscope and the LSM 5 Pascal software (Carl Zeiss, Oberkochen, Germany). Digital images were obtained at magnifications of 40× and 63×, using fluorescent filters appropriate for the specific emission wavelengths at which the secondary antibodies are visualized optimally.
In situ hybridization histochemistry. Rats treated with chronic saline or chronic amphetamine (5 mg/kg, once per day for 5 d) were decapitated 1, 3, and 14 d after the last saline or amphetamine injection. Brains were frozen in -40°C isopentane and stored at -80°C. For complimentary RNA (cRNA) generation, DNA templates for synthesizing Cav1.2 and Cav1.3 cRNA were generated from hippocampal RNA, using a PCR strategy as described previously (Kerner et al., 1998). cRNAs were targeted to a region that shows no sequence similarity between the Cav1.2 and Cav1.3 genes. Total RNA was isolated with the TRI-reagent (Sigma), and first-strand cDNA was synthesized by using the cDNA synthesis system from Invitrogen (Carlsbad, CA). Cav1.2 and Cav1.3 primers were designed by using the Oligo 4.0 software (National Biosciences, Plymouth, MN). The first-strand cDNA was used in a first round of PCR to isolate a PCR template (735 bp) that spanned both Cav1.2 and Cav1.3 sequences, using a 5′ upper primer, 5′-CTTCATCATCCTCTTCATTT-3′ (nucleotides 2950-2969 of Cav1.2, M67516; nucleotides 2703-2722 of Cav1.3, M76558), and 3′ lower primer, 5′-GCACTGGATTGAATCCCAAA-3′ (nucleotides 3704-3723 of Cav1.2 and 3418-3437 of Cav1.3). To isolate PCR templates specific for Cav1.2 and Cav1.3, we designed 40-mer inner primers (Table 1), 20 bases specific for the target sequence and 20 bases specific for the RNA polymerases SP6 (5′-GGGATTTAGGTGACACTATAGAA3-′) or T7 (5′-CTGTAATACGACTCACTATAGGG-3′). SP6 RNA polymerase was used to generate the antisense cRNA and T7 for the sense cRNA. For double ISHH in the VTA, dopamine (DA) neurons were identified by labeling TH RNA. PCR template for synthesis of the TH cRNA was generated from a TH cDNA plasmid (Table 1). In the striatum and NAc the medium spiny neurons representing the predominant cell type were identified by labeling DARPP-32 mRNA. PCR template for synthesis of the DARPP-32 cRNA was generated from striatal RNA as described above. A previously published DARPP-32 cDNA sequence (Ehrlich et al., 1990) was used for designing DARPP-32-specific PCR primers (Table 1). All primers were obtained from the Massachusetts General Hospital (MGH) DNA Oligonucleotide synthesis facility (Boston, MA). All PCR templates were sequenced by the MGH DNA Sequencing Core, Boston, MA) to confirm their identity.
Radioactive 35S-labeled Cav1.2 and Cav1.3 cRNAs were synthesized by in vitro transcription, using SP6 or T7 RNA polymerase (Promega, Madison, WI). For double in situ hybridization histochemistry (ISHH) digoxigenin (DIG)-labeled TH and DARPP-32 antisense cRNAs were synthesized with SP6 RNA polymerase and DIG 11-UTP (DIG RNA labeling mixture, Boehringer Mannheim, Indianapolis, IN).
Hybridization was performed as previously described (Kerner et al., 1998). Briefly, rats were decapitated rapidly, and brains were frozen in chilled isopentane and stored at -70°C. Then 12 μm sections were cut on a cryostat and fixed in 4% paraformaldehyde in 0.1 m PB, pH 7.4 (10 min), three washes in 0.1 m PBS, pH 7.4 (5 min each), acetylated in 0.1 m triethanolamine, pH 8.0, with 0.25% acetic anhydride (10 min), rinsed in PBS (5 min), dehydrated through 70, 80, 90, and 100% ethanol (2 min each), and delipidated in chloroform. For single ISHH 35S-labeled Cav1.2 or Cav1.3 the cRNAs were combined in a hybridization buffer containing 50% formamide, 0.3 m NaCl, 10 mm Tris, pH 8.0, 5 mm EDTA, 10% dextran sulfate, 1× Denhardt's solution, 100 mm DTT, 0.1% SDS, 0.1% sodium thiosulfate, 100 μg/ml salmon sperm DNA, 250 μg/ml yeast tRNA, and 250 μg/ml yeast total RNA. For double ISHH, 35S-labeled Cav1.2 or Cav1.3 cRNA and DIG-labeled TH cRNA or DARPP-32 cRNA were combined in the above hybridization buffer. Hybridization was performed at 50°C for 4 hr. After hybridization the sections were washed in 2× SSC at room temperature. For film autoradiograms the sections were dried and exposed to Hyperfilm β-Max film (Amersham Biosciences, Piscataway, NJ) for ∼7 d. For double ISHH the DIG-labeled TH or DARPP-32 was detected by using rabbit anti-digoxigenin antisera coupled to alkaline phosphatase (1:2000 dilution for 5 hr at room temperature; Roche Bioscience, Palo Alto, CA). Slides then were incubated in a substrate solution of nitroblue tetrazolium (0.34 mg/ml; Roche Bioscience) and bromo-chloro-indolyl-phosphate (0.18 mg/ml; Roche Bioscience) in (in mm) 100 Tris-HCl, 100 NaCl, and 50 MgCl2, pH 9.5, for 4-6 hr at room temperature. Slides were rinsed briefly in water and 70% ethanol and then dried. Slides were exposed to β-Max film (Amersham Biosciences) for ∼7 d to obtain film autoradiograms. To obtain emulsion autoradiograms, we dipped slides in Ilford K5 emulsion (Ilford Imaging, Mobberley, Cheshire, UK) diluted 1:1 with distilled water and developed after ∼3 weeks.
Quantitative analysis was performed on emulsion autoradiograms for VTA as described previously (Landwehrmeyer et al., 1995; Testa et al., 1995), using a computer-assisted M1 image analysis system (Imaging Research, St. Catherine's, Ontario, Canada). DIG-labeled neurons were visualized under bright-field optics with a 100× water immersion lens. The soma of each labeled neuron was outlined, and the overlying grains were counted via the image analysis system. Both the area of the neuronal soma, in μm2, and the number of grains present were recorded. The intensity of labeling of each neuron was computed as grains/1000 μm2. Three to four animals were used for each probe and each time point, and 25-50 neurons were counted per animal. Sections corresponding to -5.20 to -6.04 mm relative to bregma (Paxinos and Watson, 1997) were used for the analysis.
For the striatum and NAc sections the film autoradiograms were quantitated by using the M1 image analysis system calibrated with Kodak density step standards. Brain sections between 1.2 and 1.7 mm relative to bregma (Paxinos and Watson, 1997) were captured with a camera and digitized; the hybridization signal was measured as optical density (OD). Cav1.2 and Cav1.3 mRNA hybridization signal was measured in two subregions of the NAc (NAc core and NAc shell) and four striatal subregions (D, dorsal; DL, dorsolateral; DM, dorsomedial; VL, ventrolateral) as shown in Figure 6. The OD was measured in the anterior commissure and used as background level for the NAc. The OD in the corpus callosum was used as background for the striatum. Averages of the two sides for each subregion were calculated after subtraction of the background values. Four to six sections were used per animal, with three to four animals per probe and per time point.
Statistical analyses. For immunoblot, immunohistochemical analyses, and film autoradiography the data were analyzed by one-way ANOVA with post hoc comparisons (Fisher's probability of least significant difference; PLSD) between treatment groups and control. For double ISHH experiments grains/1000 μm2 for each treatment group were compared with the saline control by using ANOVA with repeated measures analysis and post hoc comparisons (Fisher's PLSD).
Acute amphetamine induces MAP kinase (ERK1/2) phosphorylation in the VTA independent of L-type Ca2+ channels
Chronic saline-treated rats were injected with saline (SAL), the LTCC antagonist diltiazem (30 and 60 mg/kg; DILT), amphetamine (5 mg/kg; AMPH), or diltiazem 20 min before an amphetamine injection (DILT+AMPH) (Table 2). Rats were decapitated 15 min after injection, and VTA tissue was isolated. Phosphorylation of ERK1/2 (P-ERK1/2) was examined by immunoblot analysis. An antibody that specifically recognizes the dually phosphorylated MAP kinase subtypes, ERK1 (P-ERK1) and ERK2 (P-ERK2) at Thr183 and Tyr185, respectively, was used. ERK1 was detected at 44 kDa and ERK2 at 42 kDa. Acute injection of amphetamine increased P-ERK1/2 in the VTA (Fig. 1A, AMPH vs SAL). The increase in P-ERK1/2 was not blocked by diltiazem (60 mg/kg) (Fig. 1A, DILT + AMPH vs AMPH). Diltiazem on its own had no effect (Fig. 1A, DILT). Similar results were obtained with 30 mg/kg diltiazem (data not shown). Blots were stripped and reprobed with an antibody that recognizes total ERK1/2. No difference in total ERK1/2 was detected across all groups. P-ERK2 bands were used for quantitation because they showed a consistent increase in phosphorylation compared with P-ERK1 in all treatment groups. P-ERK2 was normalized to total ERK2 protein and is represented as a percentage of saline control. There was a significant increase in P-ERK2 after an amphetamine challenge (Fig. 1B, A vs S; *p < 0.05), with no significant change with diltiazem pretreatment (Fig. 1B, D+A vs A; *p < 0.05 for D+A vs S).
Acute amphetamine does not induce ERK1/2 phosphorylation in chronic amphetamine-treated rats unless L-type Ca2+ channels are blocked
Chronic saline-treated rats were injected with saline (Chr SAL, SAL) or amphetamine (Chr SAL, AMPH), and chronic amphetamine-treated rats were injected with amphetamine (Chr AMPH, AMPH) or diltiazem 20 min before an amphetamine injection (Chr AMPH, DILT+AMPH) (Table 2; Fig. 2A, top). VTA tissue was isolated 15 min after injection and used for P-ERK1/2 immunoblot analysis. Compared with the P-ERK1/2 increase in chronic saline-treated rats challenged with amphetamine (Chr SAL, A vs S), there was no increase in P-ERK1/2 in chronic amphetamine-treated rats challenged with amphetamine on day 6 (Fig. 2A, Chr AMPH, A, vs Chr SAL, S). Blocking LTCCs with diltiazem (60 mg/kg) 20 min before the challenge amphetamine injection showed a significant increase in P-ERK1/2 compared with saline-treated rats (**p < 0.001; Chr AMPH, D+A, vs Chr SAL, S). Similar results were obtained with 30 mg/kg diltiazem. Chronic amphetamine-treated rats injected with either saline (Fig. 3A, Chr AMPH, SAL) or diltiazem alone (Chr AMPH, DILT+SAL) had no effect on P-ERK1/2 and looked similar to chronic SAL-treated rats (data not shown). Examination of P-ERK1/2 in the NAc, a site that receives projections from VTA dopamine neurons, revealed a significant increase (**p < 0.01) in P-ERK1/2 in chronic saline- and chronic amphetamine-treated rats after an amphetamine challenge (Fig. 2B, Chr SAL, A, and Chr AMPH, A, vs Chr SAL, S), as has been shown previously for chronic cocaine (Valjent et al., 2000). Unlike the VTA, pretreatment with diltiazem had no effect on P-ERK1/2 in the NAc (Fig. 2B, Chr AMPH, D+A vs A).
Neuroanatomic confirmation of ERK1/2 phosphorylation in the VTA
The above results were confirmed by immunohistochemical analysis with rat coronal sections containing VTA. Rats treated chronically with saline and amphetamine were challenged on day 6 as outlined in Table 2, perfused 15 min after injection, and probed with the P-ERK1/2 antibody (Fig. 3A). P-ERK1/2-positive neurons in the VTA from sections corresponding to bregma -5.20 to -6.04 mm (Paxinos and Watson, 1997) were counted (Fig. 3B). Similar to the results from immunoblot analysis, acute amphetamine significantly increased P-ERK1/2 in the VTA (**p < 0.01) (Fig. 3A,C, Chr SAL, A vs S), with no effect of diltiazem pretreatment (*p < 0.05) (Fig. 3A,C, Chr SAL, D+A vs A). There was no increase in basal P-ERK1/2 in chronic amphetamine-treated rats (Fig. 3A,C, Chr AMPH, S, vs Chr SAL, S). A challenge injection of amphetamine did not increase P-ERK1/2 in chronic amphetamine-pretreated rats (Chr AMPH, A, vs Chr SAL, S). Pretreatment with diltiazem (60 mg/kg) before the challenge amphetamine injection induced a significant increase in P-ERK1/2 (***p < 0.001; Chr AMPH, D+A, vs Chr SAL, S).
To determine whether P-ERK1/2 was expressed in VTA dopamine neurons, we performed double-label immunohistochemistry of TH and P-ERK1/2 in the Chr SAL, AMPH, and Chr AMPH, DILT+AMPH, groups (Fig. 3D). In the Chr SAL, AMPH, group P-ERK1/2-positive cells were distinct from the TH neurons (Fig. 3D, red and green arrows, respectively), whereas in the Chr AMPH, DILT+AMPH, group there was an increase in the number of P-ERK1/2-positive cells that colocalized with TH neurons (Fig. 3D, yellow arrow). In addition, a small number of P-ERK1/2-positive non-TH cells was evident (Fig. 3D, red arrow). P-ERK1/2-positive cells also were observed in the substantia nigra pars compacta (SNc) of these two groups, which did not colocalize with TH-positive neurons (data not shown).
Expression of protein phosphatases in the VTA after chronic amphetamine treatment
A decrease in P-ERK1/2 in the VTA after chronic amphetamine treatment suggested either an inactivation of the MAP kinase pathway or activation of a phosphatase pathway. We explored the second possibility by examining levels of the protein phosphatases MKP-1, MKP-3, and calcineurin (PP2B). VTA tissue was isolated from chronic saline- or amphetamine-treated (5 mg/kg) rats on day 6, 24 hr after the last injection, and was used for immunoblot analysis. There was a significant increase in MKP-1 (**p < 0.01) and PP2B (*p < 0.05) protein in chronic amphetamine-treated rats compared with chronic saline-treated rats (Fig. 4). There was no change in MKP-3 protein levels (Fig. 4).
Cav1.2 mRNA is expressed at relatively low levels in the VTA of drug-naive rats compared with Cav1.3 mRNA, which is expressed robustly
The ability of diltiazem to increase amphetamine-mediated ERK1/2 phosphorylation in chronic amphetamine-treated rats suggested that Ca2+ signaling via LTCCs was altered in these animals and that expression of one or both of the LTCC subtypes, Cav1.2 or Cav1.3, may underlie this phenomenon. To test this hypothesis, we pursued experiments to establish the normative pattern of Cav1.2 and Cav1.3 expression in the VTA, NAc, and striatum and investigated whether there were regional changes in the expression of these subtypes as a consequence of chronic amphetamine treatment.
The mRNA and protein distributions of Cav1.2 and Cav1.3 have been examined in several brain regions (Tanaka et al., 1995; Ludwig et al., 1997; Takada et al., 2001); however, the distribution in the VTA, NAc, and striatum have not been well characterized (Tanaka et al., 1995; Ludwig et al., 1997). We first examined expression in the VTA, NAc, and striatum of drug-naive rats by using in situ hybridization histochemistry (ISHH). The nucleotide sequences of Cav1.2 and Cav1.3 were compared and found to have a high degree of sequence similarity as previously reported (Ertel et al., 2000) The sequences differ over a small region that corresponds to the variable region between the second and third transmembrane domains of the Ca2+ pore-forming Cav subunit (Fig. 5A). cRNA probes specific for Cav1.2 and Cav1.3 were generated corresponding to this region and used for ISHH. ISHH was performed with 35S-labeled Cav1.2 or Cav1.3 cRNAs. In the VTA these two mRNAs were expressed differentially (Fig. 5B). Cav1.2 mRNA was present at low levels, whereas Cav1.3 mRNA was expressed strongly (Fig. 5B, top). Strong expression of both Cav1.2 and Cav1.3 was seen in the hippocampus, as previously reported (Tanaka et al., 1995; Ludwig et al., 1997). No hybridization signal was present with a sense Cav1.2 or Cav1.3 cRNA (Fig. 5B, bottom). Double ISHH was performed to examine the expression of Cav1.2 and Cav1.3 in VTA dopamine neurons. Dopamine neurons were identified with a digoxigenin (DIG)-labeled TH cRNA. Rat VTA-containing sections were hybridized with 35S-labeled Cav1.2 or Cav1.3 probes and the DIG-labeled TH cRNA. Cav1.2 and Cav1.3 mRNA was visualized as silver grains over TH-positive dopamine neurons by emulsion autoradiography. In agreement with the results from film autoradiograms, Cav1.2 mRNA was present only at low levels in VTA dopamine neurons, whereas Cav1.3 was expressed strongly (Fig. 5C).
In the NAc and striatum Cav1.2 mRNA and Cav1.3 mRNA were expressed moderately strongly, as seen by film autoradiography (Fig. 6A) and by emulsion autoradiography (Fig. 6B). A DIG-labeled DARPP-32 cRNA, a marker of medium spiny neurons of the NAc and striatum, was used to label the predominant cell type in the NAc and striatum. Cav1.2 and Cav1.3 grains were present over DARPP-32-positive cell bodies (Fig. 6B, bottom). No hybridization signal was present with the sense probes (Fig. 6A, bottom). Autoradiograms were quantitated to obtain the relative distributions of the two subtypes in subregions of the NAc and dorsal striatum. There was significantly higher expression of Cav1.2 mRNA in the NAc shell (**p < 0.01) compared with the core and the dorsal (STR D), dorsolateral (STR DL), dorsomedial (DM), and ventrolateral (VL) striatum (Fig. 6C, left). Cav1.3 was significantly higher in the NAc core (**p < 0.01) and shell (***p < 0.001) compared with the striatum (Fig. 6C, right).
Chronic amphetamine increases Cav1.2 mRNA and protein levels in the VTA, with no change in Cav1.3 mRNA or protein
To examine the expression of Cav1.2 and Cav1.3 mRNA after chronic amphetamine treatment, we decapitated chronic saline- and amphetamine-treated rats on day 6, 24 hr after the last amphetamine injection, and the brains were used for double ISHH. The density of Cav1.2 and Cav1.3 grains over TH-positive cell bodies was quantitated and is represented as grains/1000 μm2. There was a significant increase in Cav1.2 mRNA in the VTA (***p < 0.001) of amphetamine-treated rats compared with saline-treated rats (Fig. 7A). There was no significant difference in Cav1.3 mRNA between the two groups (Fig. 7C). Quantitation of grains over TH-positive neurons of the SNc showed no change in Cav1.2 or Cav1.3 mRNA (Fig. 7D,E).
Cav1.2 and Cav1.3 protein levels were examined in the VTA of chronic saline- and amphetamine-treated rats by immunoblot and immunohistochemical analyses. VTA tissue was isolated on day 6, 24 hr after the last injection. A significant increase in Cav1.2 protein was seen in chronic amphetamine-treated rats (#p < 0.0001) compared with saline-treated rats, with no change in Cav1.3 protein (Fig. 8A). To determine whether the increase in Cav1.2 protein was present in TH-positive neurons, we used double labeling. TH neurons were labeled with a green fluorescent label and Cav1.2 with a red fluorescent label. An increase in Cav1.2 protein was seen over TH-positive neurons after chronic amphetamine treatment (Fig. 8B, compare left and right) although such induction of Cav1.2 protein was additionally evident in non-TH-containing neurons (Fig. 8B, right, white arrow).
To examine further whether the increase in P-ERK1/2 seen in the Chr AMPH, DILT+AMPH, group (Fig. 3A,D) was present in neurons that exhibited an increase in Cav1.2 protein, we used double labeling (Fig. 8C). P-ERK1/2 was labeled with a green fluorescent label and Cav1.2 with a red fluorescent label. P-ERK1/2-positive neurons colocalized with the neurons that had an increase in Cav1.2 protein [compare P-ERK1/2 alone with green labeling (left), Cav1.2 alone with red labeling (middle), and the merged image of P-ERK1/2 and Cav1.2 (right)].
Increase in Cav1.2 mRNA expression is maintained in the VTA at 3 and 14 d of withdrawal
Chronic amphetamine treatment results in molecular and behavioral changes that are maintained even after long periods of withdrawal (Wolf, 1998). Hence we examined the expression of Cav1.2 and Cav1.3 at 3 and 14 d after the last amphetamine injection in VTA dopamine neurons. Cav1.2 mRNA levels were significantly higher in chronic amphetamine-treated compared with chronic saline-treated rats at 3 and 14 d of withdrawal (Fig. 9). However, there was a trend toward a decrease in the fold induction of Cav1.2 mRNA levels from 1 to 3 to 14 d of withdrawal (Figs. 7A, 9A). There was no change in Cav1.3 mRNA at 3 and 14 d of withdrawal (Fig. 9B).
Cav1.2 mRNA is increased in the nucleus accumbens for 1, but not for 3 or 14, d after chronic amphetamine treatment
Autoradiograms from ISHH with chronic saline- and amphetamine-treated rat sections containing the NAc and striatum (between 1.0 and 2.0 mm anterior to bregma; Paxinos and Watson, 1997) at 1, 3, and 14 d were used for quantitation. There was a significant increase (*p < 0.05) in Cav1.2 mRNA in the NAc shell 1 d after chronic amphetamine treatment that returned to basal levels at 3 and 14 d of withdrawal (Fig. 10A). There was no change in Cav1.2 mRNA expression in the NAc core or of Cav1.3 mRNA in the NAc shell or core (Fig. 10A,B). No change in Cav1.2 or Cav1.3 mRNA was present in the subregions of the striatum after chronic amphetamine treatment (data not shown).
We report here for the first time a study that has examined the role of LTCCs in amphetamine-mediated molecular changes in the VTA. We find that LTCCs do not play a role in acute amphetamine-mediated ERK1/2 phosphorylation in the VTA. In chronic amphetamine-treated rats we find no increase in amphetamine-mediated ERK1/2 phosphorylation unless LTCCs are blocked, in which case there is robust phosphorylation of ERK1/2 in dopaminergic VTA neurons. We also find that chronic amphetamine treatment is accompanied by an increase in the LTCC subtype Cav1.2 mRNA and protein in dopaminergic VTA neurons and in the phosphatases calcineurin (PP2B) and MKP-1. Together, our studies identify that chronic amphetamine upregulates Cav1.2-containing LTCCs, which activate a phosphatase pathway that results in decreased ERK1/2 phosphorylation in the dopaminergic neurons of the VTA (Fig. 11).
ERK1/2 pathway activation in the VTA after acute versus chronic psychostimulants
Our finding of an increase in ERK1/2 phosphorylation after acute amphetamine-treated, but not after chronic amphetamine-treated, rats was unexpected because the opposite has been reported for cocaine (Berhow et al., 1996). The difference in our results may be attributable to the different drugs that were used or to the different time points that were examined. For their acute experiments Berhow et al. (1996) measured ERK1/2 activity 2 hr after cocaine injection, a time by which phosphorylation already may have returned to basal levels. Our studies were performed 15 min after amphetamine injection. In their chronic studies Berhow et al. (1996) reported an increase in basal ERK1/2 activity in the VTA 10 hr after chronic cocaine administration, although we found no basal P-ERK1/2 or amphetamine-induced P-ERK1/2 when the VTA was examined 24 hr after the last amphetamine injection. Another reason for the discrepant results may be attributable to the different assays used; Berhow et al. (1996) studied ERK1/2 activity by using biochemical assays, whereas we detected ERK1/2 phosphorylation by immunoblotting and immunohistochemistry.
Two possible routes of ERK1/2 phosphorylation in the VTA are via the NMDA receptor or the neurotrophic factor neurotrophin-3. NMDA receptors are essential for psychostimulant-induced Ca2+ signaling (Kalivas and Alesdatter, 1993; Churchill et al., 1999; Pierce et al., 1999; Vezina and Queen, 2000; Ungless et al., 2001; Suto et al., 2003) and activate ERK1/2 phosphorylation (Sweatt, 2001). Neurotrophin-3 via the TrkC receptors also activates the ERK1/2 pathway in the VTA (Pierce et al., 1999). Consistent with this is the fact that acute amphetamine has been found to increase calmodulin (CaM) mRNA and protein in the VTA (Michelhaugh and Gnegy, 2000) that would activate the ERK1/2 pathway.
The role of Cav1.2 in ERK1/2 pathway activation in the VTA after acute versus chronic psychostimulants
The absence of a role of LTCCs in acute amphetamine-mediated ERK1/2 correlated with the presence of very low levels of the LTCC subtype Cav1.2 mRNA and protein in the VTA of drug-naive rats. The involvement of LTCCs in regulating chronic amphetamine-mediated ERK1/2 phosphorylation corresponds with an increase in Cav1.2 mRNA and protein. The Cav1.2 gene contains the upstream regulatory sites CRE and AP1, which bind CREB and the Fos family of proteins, respectively (Liu et al., 2000). These regulatory sites and their transcription factors are activated by psychostimulants and are involved in psychostimulant-induced changes (Berke and Hyman, 2000; Nestler, 2001a). Because chronic amphetamine-induced long-lasting changes require new protein synthesis (Karler et al., 1993; Sorg and Ulibarri, 1995), the increase in Cav1.2 mRNA may contribute to psychostimulant-induced neuroplasticity. Chronic nicotine (Katsura et al., 2002), ethanol (Walter and Messing, 1999; Walter et al., 2000), and norepinephrine (Maki et al., 1996; Golden et al., 2002) also have been found to increase the expression of Cav1.2.
Activation of phosphatases in the VTA after acute versus chronic psychostimulants
In chronic amphetamine-treated rats we find that LTCCs downregulate ERK1/2 phosphorylation. LTCCs are a major source of activation of Ca2+/CaM kinases but also have been found to activate the phosphatase pathway (Groth et al., 2003). We find that in chronic amphetamine-treated rats there is an increase in the phosphatase calcineurin (PP2B). This is in line with other studies that have shown an increase in calcineurin activity via LTCCs (Graef et al., 1999; Chang and Berg, 2001; Snyder et al., 2003). In addition, acute amphetamine has been found to increase PP2B expression (Wang and Uhl, 1998).
We also find an increase in MKP-1, a phosphatase that directly dephosphorylates ERK1/2 at the Tyr and Thr residues. It has been found that ERK1/2 and MKP functions are tightly linked. P-ERK1/2 has been found to phosphorylate MKP, rendering the protein more resistant to degradation (Brondello et al., 1999) and thus creating a negative feedback loop for its own phosphorylation (Bhalla et al., 2002). In addition, PP2B has been found to increase MKP-1 mRNA expression (Lim et al., 2001). In response to psychostimulants, MKP-1 mRNA expression has been found to increase after acute cocaine and methamphetamine (Berke et al., 1998; Takaki et al., 2001; Ujike et al., 2002) and to persist after chronic methamphetamine treatment (Takaki et al., 2001; Ujike et al., 2002). Our results suggest that acute amphetamine increases MKP-1 expression via an increase in ERK1/2 phosphorylation and in PP2B expression. The increased levels of MKP-1 and PP2B then would downregulate the phosphorylation of ERK1/2.
The role of LTCCs in mediating psychostimulant-induced neuroadaptations
A lack for a role of LTCCs in ERK1/2 phosphorylation in the NAc was surprising. LTCCs mediate dopamine and glutamate signaling in the NAc and striatum (Surmeier et al., 1995; Liu and Graybiel, 1996, 1998; Cepeda et al., 1998; Rajadhyaksha et al., 1999). One possible explanation could be a role for LTCCs in long-term ERK1/2 phosphorylation similar to what has been seen in the hippocampal and cortical cultures (Bito et al., 1996; Dolmetsch et al., 2001). This would be consistent with a role of LTCCs and the MAP kinase pathway in activation of long-term P-ERK1/2 and gene expression (Dolmetsch et al., 2001; Wu et al., 2001a,b), which may contribute to long-term psychostimulant-induced neuronal plasticity.
The LTCC-mediated decrease in ERK1/2 phosphorylation may represent a homeostatic adaptation after chronic amphetamine exposure. The MAP kinase pathway specifically via activation of MKPs has been suggested as one route that neurons use in adaptive responses to chronic stimuli (Bhalla et al., 2002). A decrease in ERK1/2 phosphorylation has been reported after chronic ethanol exposure, which becomes even more robust with an increase in withdrawal times (Sanna et al., 2002). Further examination of the state of ERK1/2 phosphorylation after longer periods of withdrawal will address this question.
The role of LTCCs in mediating long-term potentiation
Recent studies have shown that a single exposure to amphetamine or cocaine induces long-term potentiation (LTP) at the glutamatergic synapse on VTA dopamine neurons (Ungless et al., 2001; Saal et al., 2003) and is believed by some investigators to be a physiological correlate of psychostimulant-induced sensitized behavior. The same synapse has been shown to express voltagegated Ca2+ channel-dependent long-term depression (LTD) that is inhibited by amphetamine (Jones et al., 2000). LTCCs contribute to an NMDA-independent form of LTP (Johnston et al., 1992), and in light of our current findings we hypothesize that Ca2+ influx from Cav1.3-containing LTCCs that are expressed predominantly in VTA dopamine neurons may contribute to LTD and that the increase in Cav1.2 expression after amphetamine treatment may mediate an LTCC-dependent LTP. This is in line with the localization of Cav1.2-containing LTCCs at synapses (Hell et al., 1993; Furuyashiki et al., 2002) and several reports showing a role for LTCCs in neuroadaptive changes in the brain (Weisskopf et al., 1999; Bauer et al., 2002; Deisseroth et al., 2003).
The influence of LTCC on the ERK1/2 pathway and psychostimulant-induced behavioral sensitization
Our finding of a role for LTCCs in chronic, but not acute, amphetamine-mediated ERK1/2 phosphorylation also agrees with psychostimulant-induced behavioral data. It has been reported that LTCC blockers do not block the acute effects of amphetamine and cocaine but block chronic amphetamine- and cocaine-induced locomotor activity (Karler et al., 1991; Pierce and Kalivas, 1997; Pierce et al., 1998). It is interesting that inhibiting the MAP kinase pathway does not block acute cocaine-induced locomotor activity but inhibits such effects after chronic cocaine administration (Pierce et al., 1999). This suggests that the LTCC-activated MAPK pathway may contribute to the adaptations specific to chronic psychostimulant exposure. In addition, we find that Cav1.2 mRNA is upregulated up to 14 d of withdrawal. LTCCs have been suggested to play a role in the augmented dopamine release observed during the expression of sensitized behaviors (Pierce and Kalivas, 1997). Although the neuronal distribution and physiology of Cav1.2 do not support a role in neurotransmitter release, LTCCs have been found to be involved in neurotransmitter release (Watanabe et al., 1998; Evans and Pocock, 1999; Okita et al., 2000), and Cav1.2 recently has been suggested to play this role in fear-mediated behavior (Shinnick-Gallagher et al., 2003). Additional experiments are required to explore further the relevance of this mechanism in contributing to long-term changes in neurotransmitter release in forebrain targets innervated by VTA neurons, which may contribute to recurrent psychostimulant-induced behaviors.
This work has been supported by National Institute on Drug Abuse (NIDA) Grants KO1DA14057 (A.R.) and K02DA00354 (B.K.) and a NIDA International Visiting Scientists and Technical Exchange Program (INVEST) fellowship (I.H.). We thank Dr. C. J. Malanga for his advice and help in the preparation of this manuscript, Jennifer Johnson and Dr. David Miller for their technical assistance with the in situ hybridization histochemistry experiments, Deirdre McCarthy and Dr. Pradeep Bhide for their assistance with the double immunohistochemistry experiments, and Igor Bagayev for his assistance with confocal microscopy.
Correspondence should be addressed to Anjali Rajadhyaksha, Molecular and Developmental Neuroscience, Room 2415, NMR Center, Department of Radiology, Massachusetts General Hospital, Harvard Medical School, Building 149, Thirteenth Street, Charlestown, MA 02129. E-mail:.
Copyright © 2004 Society for Neuroscience 0270-6474/04/249464-13$15.00/0