Cocaine-induced CREB phosphorylation in nucleus accumbens of cocaine-sensitized rats is enabled by enhanced activation of extracellular signal-related kinase, but not protein kinase A

Authors

  • Brandi J. Mattson,

    Search for more papers by this author
    • 1

      The present address of Brandi J. Mattson is Laboratory of Neocortical Microcircuitry, Brain and Mind Institute, Ecole Polytechnique Federale de Lausanne, Bat AA-B, Station 15, Lausanne 1015, Switzerland.

  • Jennifer M. Bossert,

    1. Behavioral Neuroscience Branch, Intramural Research Program, National Institute on Drug Abuse, National Institutes of Health, Department of Health and Human Services, Baltimore, Maryland, USA
    Search for more papers by this author
  • Danielle E. Simmons,

    1. Behavioral Neuroscience Branch, Intramural Research Program, National Institute on Drug Abuse, National Institutes of Health, Department of Health and Human Services, Baltimore, Maryland, USA
    Search for more papers by this author
  • Naohito Nozaki,

    Search for more papers by this author
    • 2

      The present address of Naohito Nozaki is Department of Biochemistry and Molecular Biology, Kanagawa Dental College, 82 Inaoka-cho, Yokosuka, Kanagawa, 238–8580, Japan.

  • Deepti Nagarkar,

    1. Behavioral Neuroscience Branch, Intramural Research Program, National Institute on Drug Abuse, National Institutes of Health, Department of Health and Human Services, Baltimore, Maryland, USA
    Search for more papers by this author
  • Justin D. Kreuter,

    1. Behavioral Neuroscience Branch, Intramural Research Program, National Institute on Drug Abuse, National Institutes of Health, Department of Health and Human Services, Baltimore, Maryland, USA
    Search for more papers by this author
  • Bruce T. Hope

    1. Behavioral Neuroscience Branch, Intramural Research Program, National Institute on Drug Abuse, National Institutes of Health, Department of Health and Human Services, Baltimore, Maryland, USA
    Search for more papers by this author

Address correspondence and reprint requests to Bruce T. Hope, PhD, National Institute on Drug Abuse, Behavioral Neuroscience Branch, 5500 Nathan Shock Drive, Building C, Baltimore, MD 21224, USA. E-mail: bhope@intra.nida.nih.gov

Abstract

Repeated cocaine administration to rats outside their home cages sensitizes the behavioral effects of the drug, and enhances induction of the immediate early gene product Fos in nucleus accumbens. We hypothesized that the same treatment regimen would also enhance cocaine-induced activation of intracellular signaling kinases that phosphorylate cyclic AMP-regulated element-binding protein (CREB), an important mediator of c-fos transcription. Phosphorylation levels of extracellular signal-regulated kinase (ERK)/mitogen-activated protein kinase (MAPK), calcium/calmodulin kinases (CaMKs) II and IV, and CREB were used to assess endogenous functional activity of these signaling molecules in rats behaviorally sensitized outside their home cages. Protein kinase A (PKA)-specific phosphorylation of Ser845 in the α-amino-3-hydroxy-5-methylisoxazole-4-propionate (AMPA) receptor subunit GluR1 was used to assess endogenous functional activity of PKA. Using western blots and immunohistochemistry, we detected cocaine-induced CREB phosphorylation after repeated cocaine administration, but not after repeated saline administration. Using western blots and MAPK activity assays, we found that cocaine-induced phosphorylation and activation of ERK, but not of CaMKs II or IV or GluR1, was augmented in nucleus accumbens of cocaine-sensitized rats. Unilateral infusions of the MAPK kinase inhibitor U0126 into nucleus accumbens attenuated cocaine-induced ERK and CREB phosphorylation in cocaine-sensitized rats. In contrast, unilateral infusions of the PKA inhibitor Rp-isomer of adenosine-3′,5′-cyclicmonophosphorothioate (Rp-cAMPs) did not affect cocaine-induced CREB phosphorylation. Therefore, enhanced activation of ERK, but not PKA, enables and mediates cocaine-induced CREB phosphorylation in nucleus accumbens of rats that are sensitized by repeated cocaine administration outside their home cages.

Abbreviations used
ACSF

artificial cerebrospinal fuid

BSA

bovine serum albumin

CaMK

calcium/calmodulin kinase

cAMP

cyclic AMP

CREB

cAMP-regulated element-binding protein

DMSO

dimethylsulfoxide

ERK

extracellular signal-regulated kinase

GluR1

glutamate receptor subunit type-1

MAPK

mitogen-activated protein kinase

MEK

MAPK kinase

PBS

phosphate-buffered saline

PKA

protein kinase A

Rp-cAMPs

Rp-isomer of adenosine-3′,5′-cyclicmonophosphorothioate

SDS

sodium dodecyl sulfate

Cocaine increases locomotor activity and behavioral stereotypies in rats (Vanderschuren and Kalivas 2000). Rats become sensitized to these drug actions following repeated cocaine administration, particularly when the drug is repeatedly administered in an environment other than the rat's home cage. The effects of cocaine in the striatum are clearly implicated in these behavioral responses (Kelly and Iversen 1976; Delfs et al. 1990), and many investigators have focused on identifying neuroadaptations in this brain region that might underlie altered responsiveness in sensitized animals (Vanderschuren and Kalivas 2000; Hyman and Malenka 2001; Nestler 2001).

One such neuroadaptation is altered phosphorylation of the transcription factor cyclic AMP (cAMP)-regulated element-binding protein (CREB) in the striatum. Phosphorylation of CREB on Ser133 activates CREB and increases transcription of a large number of genes that can alter neuronal function and regulate synaptic plasticity (Frank and Greenberg 1994; Mayr and Montminy 2001). Drug-induced CREB phosphorylation in the striatum is altered 1 day after repeated injections of high doses of the cocaine-related psychostimulant amphetamine in the home cage (Cole et al. 1995; Simpson et al. 1995; Turgeon et al. 1997). However, CREB phosphorylation has not been examined after repeated cocaine administration or after longer drug withdrawal periods, even though cocaine-sensitized locomotor activity persists for at least 30 days (Henry and White 1995).

CREB phosphorylation has not been examined following repeated administration of cocaine or amphetamine outside the rat's home cage. Repeated cocaine administration outside the home cage effects sensitized behavior and striatal Fos (an important target gene of CREB) induction quite differently from repeated home-cage administration (Vanderschuren and Kalivas 2000; Crombag et al. 2002; Todtenkopf et al. 2002; Badiani and Robinson 2004). These findings suggest that the mechanisms mediating CREB phosphorylation might also be affected by environmental factors associated with repeated administration of cocaine.

CREB can be phosphorylated by several different kinase pathways (Shaywitz and Greenberg 1999; Mayr and Montminy 2001), including cyclic AMP-dependent protein kinase (PKA) (Montminy and Bilezikjian 1987). Neural activity can also induce CREB phosphorylation by increasing calcium influx that activates calcium/calmodulin-dependent kinases (CaMKs I, II and IV) (Sheng et al. 1991; Enslen et al. 1994; Matthews et al. 1994; Braun and Schulman 1995; Tokumitsu and Soderling 1996; Heist et al. 1998; Tokumitsu et al. 2004), and extracellular signal-regulated kinase (ERK) (Xing et al. 1996; Sgambato et al. 1998; Thomas and Huganir 2004).

Very little is known about how cocaine-induced activation of these kinases is altered by repeated cocaine administration, particularly when the drug is administered outside the rat's home cage. Although it is well established that these kinases are capable of phosphorylating CREB in vitro and in vivo, it is not known which kinase is responsible for altering levels of cocaine-induced CREB phosphorylation in nucleus accumbens of cocaine-sensitized rats. In the present study, we assessed indicators of the levels of endogenous activity of these kinases and CREB phosphorylation in the nucleus accumbens and caudate–putamen of rats 7 days after repeated cocaine administration outside their home cages. We then explored a causal role for enhanced cocaine-induced ERK activation in mediating cocaine-induced CREB phosphorylation in the nucleus accumbens of conscious behaving rats.

Materials and methods

Subjects

Male Sprague-Dawley rats (200–250 g; n = 275) from Charles River (Raleigh, NC, USA) were housed individually and acclimatized to a reverse 12 : 12 h light–dark cycle (lights off at 07.00 hours) for a minimum of 5 days before the start of experiments. Rats were given free access to food and water. The animal facility was accredited through the American Association for Accreditation of Laboratory Animal Care. Experimental procedures followed the guidelines described in Principles of Laboratory Animal Care (National Institutes of Health publication no. 86-23, 1996).

Locomotor activity

Cocaine was administered repeatedly to rats to produce sensitization of locomotor activity (Crombag et al. 2002; Hope et al. 2005). Rats were injected intraperitoneally once daily for 7 days with cocaine (15 mg/kg) or saline vehicle (1 mL/kg) in locomotor activity chambers (Medical Associates, Georgia, VT, USA). For each injection, rats were brought from their home cages and habituated to the activity chambers for 30 min. Rats were then injected with cocaine or saline and their activity was monitored for 60 min. Rats were kept in their home cages for 7 days after the last cocaine injection. In the first set of experiments, the time course for protein phosphorylation was examined and all rats received a challenge intraperitoneal injection of 15 mg/kg cocaine. The second set of experiments used a 2 × 2 treatment regimen whereby each of the groups that received repeated cocaine or saline was further divided into two groups that received a challenge injection of 15 mg/kg cocaine or saline in the activity chambers. This dose is commonly used to assess sensitized locomotor activity in behavioral studies; we avoided higher cocaine doses like those commonly used in molecular studies because they might activate molecular or cellular mechanisms unrelated to sensitized locomotor activity.

Western blotting

Rats were killed by rapid decapitation 0, 10, 20, 40 and 60 min after challenge injections. Their brains were extracted and frozen in isopentane (− 50°C) within 30–45 s of decapitation and stored at − 80°C until dissection. One-millimeter coronal slices of brain were cut in a cryostat (Reichert-Jung 2800E Leica Inc., Deerfield, IL, USA) kept at − 20°C. Tissue punches were obtained from caudate–putamen (12 G) and nucleus accumbens (14 G) (see Fig. 1) and then sonicated in 1% sodium dodecyl sulfate (SDS). Tissue was kept frozen at all times until sonication in SDS. Protein concentrations of the samples were determined using the bicinchoninic acid assay (Pierce Chemical Company; Rockford, IL, USA). Sample concentrations were equalized by diluting with 1% SDS.

Figure 1.

Schematic representation of the regions of nucleus accumbens (Acb) and caudate–putamen (CPu) where tissue punches were obtained for molecular assays. The circles demarcate the regions obtained in the tissue punches. Only the rostral face (+ 1.8 mm from Bregma) of the coronal section is shown. The circles also depict approximations of the regions used for analyses of phospho-CREB immunohistochemistry. Stereotaxic co-ordinates and drawings are adapted from the atlas of Paxinos and Watson (1998). ac, anterior commisure; cc, corpus callosum.

Samples were subjected to SDS–polyacrylamide gel electrophoresis (10% acrylamide/0.27%N,N′-methylenebisacryalamide resolving gel) for 3 h at 200 V. For each electrophoresis run, increasing amounts of protein pooled from the brain region being tested were run alongside the individual samples and used to produce a standard curve. Proteins were transferred electrophoretically to Immobilon-P membranes (Millipore Corp; Bedford, MA, USA) at 0.3 A for 2 h. Membranes were incubated four times each for 15 min in blocking buffer containing 2% non-fat dry milk in PBST [10 mm phosphate-buffered saline (PBS) plus 0.05% Tween-20] for phospho-ERK, ERK, phospho-CaMKII, CaMKII, phospho-glutamate receptor subunit type-1 (GluR1) and GluR1; 1% non-fat dry milk in PBST for phospho-CREB; 0.5% non-fat dry milk in PBST for phospho-CaMKIV; and 1% polyvinyl-pyrollidone in PBST for CREB and CaMKIV. Membranes were then incubated overnight at 4°C with primary antibody diluted in their respective blocking buffers. The antibodies used were anti-phospho-CaMKII (1 : 2000 dilution; Upstate Biotechnology, Lake Placid, NY, USA), anti-CaMKII (1 : 10 000 dilution; Upstate Biotechnology), anti-phospho-CaMKIV (1 : 10 dilution, mouse monoclonal antibody; Tokumitsu et al. 2004), anti-CaMKIV (1 : 1000 dilution; BD Biosciences Pharmingen, San Diego, CA, USA), anti-phospho-ERK (1 : 8000 dilution; Cell Signaling, Beverly, MA, USA), anti-ERK (1: 2000 dilution; Cell Signaling), anti-phospho-CREB (1 : 1000 dilution; Upstate Biotechnology), anti-CREB (1 : 1000 dilution; Cell Signaling), anti-phospho-GluR1 (phospho-Ser845; 1 : 2000 dilution; Upstate Biotechnology) and anti-GluR1 (1 : 5000 dilution; Chemicon International, Temecula, CA, USA).

After incubation with primary antibodies, blots were processed with peroxidase-labeled secondary antibodies (Vector Laboratories, Burlingame, CA, USA) and detected using the enhanced chemiluminescence (ECL) procedure of Amersham Biosciences (Piscataway, NJ, USA). Luminescence from the blots was detected using ECL Hyperfilm (Amersham Biosciences) followed by digital scanning in transparency mode. Band intensities were quantified using Quantity One software (Version 4.3; Biorad Corporation, Hercules, CA, USA). Band intensities from test samples were compared with band intensities from the standard curve. The amount of the protein of interest in each sample was interpolated from the standard curves run with all western blots to ensure that samples were within the linear range of detection. For each protein, phosphorylated protein levels were normalized to total protein levels to determine the phosphorylation state within each brain region.

The amounts of sample protein left in each sample for assays of CaMKIV phosphorylation were very low. Thus only three to four samples were assayed for each group from the 20-min post-challenge experiment and total CaMKIV levels were not assayed for the 10- or 20-min post-challenge experiments.

MAPK activity assays

Tissue punches were obtained from nucleus accumbens (14 G) and sonicated in 300 µL lysis buffer (20 mm Tris, 150 mm NaCl, 1 mm EDTA, 1 mm EGTA, 1% Triton X-100, 2.5 mm sodium pyrophosphate, 1 mmβ-glycerolphosphate, 1 mm Na3VO4, 1 µg/mL leupeptin and 1 mm phenylmethylsulfonyl fluoride). Samples were centrifuged at 20 000 g for 10 min at 4°C. The supernatants were removed and assayed for protein concentration using the bicinchoninic acid assay (Pierce Chemical Company). Samples were diluted to 1 mg/mL with lysis buffer and assayed for kinase activity using an MAPK assay kit (Cell Signaling). Two hundred microliters of cell lysate (200 µg total protein) from each sample was processed according to the kit manual. Elk-1 fusion protein was used as the substrate and the phosphorylated Elk-1 fusion protein product was analyzed using the western blotting technique described above.

Immunohistochemistry for phosphorylated CREB

To examine phospho-CREB immunoreactivity, rats were deeply anesthetized with equithesin 20 min after challenge injections of cocaine or saline. Rats were then perfused transcardially with 100 mL 0.1 m PBS, pH 7.4, followed by 300 mL 4% paraformaldehyde in 0.1 m sodium phosphate, pH 7.4. Brains were removed and post-fixed for 2 h before being transferred to 20% sucrose in 0.1 m sodium phosphate, pH 7.4, for 48 h at 4°C. Brains were frozen in powdered dry ice and stored at − 30°C. Coronal sections (30 µm) were cut in a cryostat, collected in cryoprotectant [20% glycerol and 2% dimethylsulfoxide (DMSO) in 0.1 m sodium phosphate, pH 7.4], and stored at − 80°C until further processing.

Sections were rinsed four times each for 10 min in PBS and placed in blocking buffer for 1 h at 22°C. Blocking buffer was 3% bovine serum albumin (BSA), 0.3% Triton X-100 in PBS for phospho-CREB immunohistochemistry. Sections were incubated overnight at 4°C in a 1 : 500 dilution of anti-phospho-CREB antibody (Cell Signaling) in PBS with 0.3% Triton X-100 and 3% BSA. Sections were washed in PBS and incubated with biotinylated goat anti-rabbit IgG secondary antibody (Vector Laboratories) at a dilution of 1 : 500 in 3% BSA, 0.3% Triton X-100 in PBS. Sections were washed again in PBS and processed with ABC Elite kits (Vector Laboratories). Sections were developed in a solution containing 0.035% diaminobenzidine, 2.5% nickel ammonium sulfate (for enhanced visualization of the chromagen) and 0.01% hydrogen peroxide in 100 mm sodium acetate buffer, pH 7.0.

Bright-field images of phosphorylated CREB immunoreactivity in caudate–putamen and nucleus accumbens (1.8 mm anterior to Bregma; Paxinos and Watson 1998) were captured using a charge-coupled device camera (Coolsnap Photometrics; Roper Scientific Inc., Trenton, NJ, USA) attached to a Axioskop 2 microscope Carl Zeiss MicroImaging Inc., Thornwood, NY, USA. The size of the image captured was 1.76 × 1.36 mm, with a total area of 2.39 mm2. Labeled nuclei in two hemispheres per rat were automatically counted using IPLab software (Scanalytics, Inc., Fairfax, VA, USA).

Surgery and intra-accumbens infusions of Rp-cAMPs and U0126

Three different experiments examined the effects of infusions of the PKA inhibitor Rp-cAMPs and the MAPK kinase (MEK) inhibitor U0126 on CREB phosphorylation in the nucleus accumbens. Forty-eight rats (250–300 g) were anesthetized with sodium pentobarbital + chloral hydrate (60 + 225 mg/kg; i.p.). Permanent guide cannulae (23 G; Plastics One, Roanoke, VA, USA) were implanted bilaterally at a 10° angle with the tip 1 mm above the nucleus accumbens; coordinates were anteroposterior + 1.6 mm, mediolateral ± 1.6 mm, and dorsoventral − 6.5 mm relative to Bregma (Paxinos and Watson 1998). The cannulae were fixed in place with acrylic dental cement and secured by jeweler's screws, and stylets were placed in the guide cannulae to prevent clogging. Rats were then kept in their home cages for at least 5 days before drug treatments.

Intracranial experiment 1: U0126 infusions and repeated drug administration

Six rats were repeatedly administered 15 mg/kg cocaine and then kept in their home cages for 7 days as described above. On test day, 1 µg U0126 (Calbiochem, La Jolla, CA, USA) in 0.5 µL was infused unilaterally into the nucleus accumbens of one hemisphere and 0.5 µL vehicle was infused into the other. U0126 was dissolved first in DMSO to 4 µg/µL and diluted to 2 µg/µL with artificial cerebrospinal fluid (ACSF; 148 mm NaCl, 2.7 mm KCl, 1.2 mm CaCl2, 0.8 mm MgCl2, pH 7.4); vehicle infusions were 50% DMSO in ACSF. Intracranial infusions were performed using a syringe pump (Harvard Apparatus Inc., Holliston, MA, USA) and 10-µL Hamilton syringes (Hamilton Co., Reno, NV, USA) connected via polyethylene-50 tubing to 30-G injectors (Plastics One). Infusions were performed over 1 min and the injectors were left in place for an additional 1 min, followed by re-insertion of the stylets. The hemispheres receiving drug or vehicle infusions were counterbalanced. Rats were placed in locomotor activity chambers immediately after the intracranial infusions and challenge injections of 15 mg/kg cocaine were administered 30 min after infusions.

Intracranial experiment 2: U0126 infusions and challenge injections

Twelve rats were repeatedly administered 15 mg/kg cocaine and then kept in their home cages for 7 days as described above. On test day, 1 µg U0126 in 0.5 µL was infused unilaterally into the nucleus accumbens of one hemisphere and 0.5 µL vehicle was infused into the other. Rats were placed in locomotor activity chambers immediately after the intracranial infusions. Challenge injections of 15 mg/kg cocaine or saline (n = 6 per group) were administered 30 min after infusions.

Intracranial experiment 3: Rp-cAMPs infusions

Six rats were repeatedly administered 15 mg/kg cocaine and then kept in their home cages for 7 days as described above. On test day, 80 nmol Rp-cAMPs (Calbiochem) in 0.5 µL was infused unilaterally into the nucleus accumbens of one hemisphere and 0.5 µL vehicle was infused into the other. Rp-cAMPs was dissolved in ACSF to a concentration of 160 nmol/µL; vehicle infusions were ACSF. Rats were placed in locomotor activity chambers immediately after the intracranial infusions and challenge injections of 15 mg/kg cocaine were administered 30 min after infusions.

Processing

For all intracranial experiments, rats were decapitated 20 min after challenge injections and their brains were extracted and frozen in isopentane as above. Tissue punches of nucleus accumbens were obtained and assayed for phospho-CREB, CREB, phospho-ERK, ERK, phospho-GluR1, GluR1, phospho-CaMKIV and CaMKIV by western blotting.

Statistical analysis

Bartlett's test for homogeneity of variance was applied before performing two-way factorial anova followed by Fisher's PLSD (Protected Least Significant Difference) post hoc tests to examine the main effects and their interactions. All tests were performed using a significance level of p < 0.05. For all western blot and enzyme activity experiments, statistical analyses were performed on data before transforming them to percentage of saline control.

Results

Cocaine-induced locomotor activity

Repeated cocaine administration significantly enhanced levels of cocaine-induced locomotor activity in all 2 × 2 factorial design experiments whether rats were killed 10 or 20 min following challenge injections (for example, see Fig. 2). In all experiments, two-way anovas showed significant interactions between repeated administration and challenge drug, and post hoc comparisons indicated that cocaine challenge injections induced more locomotor activity after repeated cocaine administration. Locomotor activity levels following saline challenge injections were similar in rats repeatedly administered cocaine or saline, indicating that repeated cocaine administration did not produce conditioned locomotor activity.

Figure 2.

Cocaine-induced locomotor activity is enhanced by repeated cocaine administration. Values are mean ± SEM (n = 8 per group) of the distance traveled following repeated administration of cocaine or saline in response to a challenge injection of cocaine or saline. All groups were analyzed using 2-way anova. Fishers PLSD was used for post-hoc comparisons. †Indicates significantly more distance traveled compared with levels following saline challenge injections. ‡Indicates significantly more distance traveled compared with saline challenge injections or following cocaine challenge injections in rats repeatedly administered saline.

Cocaine-induced CREB phosphorylation

We assessed the phosphorylation state of Ser133 in CREB in the nucleus accumbens and caudate–putamen of rats repeatedly administered cocaine or saline at 0, 10, 20, 40 and 60 min following challenge injections of cocaine. This time course was a preliminary examination of phosphorylation kinetics to determine times for peak differences in CREB phosphorylation following repeated administration of cocaine versus repeated administration of saline. In all cases, CREB phosphorylation levels were calculated as the ratio of Ser133 phosphorylation to total CREB protein levels. Total CREB protein levels were not significantly altered in any of our experiments (data not shown). The focus of our study was regulation in the nucleus accumbens because of its relevance to cocaine's locomotor and rewarding effects. However, we added the caudate–putamen for comparison in our initial experiments because of previous findings that repeated cocaine administration outside the rats' home cage enhances cocaine-induced Fos expression in nucleus accumbens but not in the caudate–putamen.

Nucleus accumbens

In the time course experiment, repeated cocaine administration significantly altered cocaine-induced CREB phosphorylation levels in the nucleus accumbens (two-way anova, effect of repeated administration, F1,56 = 5.13, p < 0.05; effect of time, F4,56 = 2.60, p < 0.05; repeated administration × time, F4,56 = 3.82, p < 0.05) (Fig. 3a). Post hoc comparisons indicated that CREB phosphorylation levels were significantly higher at 10 and 20 min, but not at 40 and 60 min, in rats repeatedly administered cocaine relative to levels in rats repeatedly administered saline.

Figure 3.

Cocaine-induced CREB phosphorylation is enhanced in the nucleus accumbens following repeated cocaine administration. Time course is shown for cocaine-induced CREB phosphorylation following repeated administration of cocaine or saline in (a) nucleus accumbens and (d) caudate–putamen. Values are expressed as mean ± SEM (n = 6–8 per group) transformed to percentage of control values (repeated saline/0-min control). ‡Indicates significantly higher levels at 10 and 20 min following repeated cocaine administration compared with repeated saline administration. Blot insets show representative bands from western blots of cocaine-induced phosphorylated CREB following repeated saline administration (S) and repeated cocaine administration (C). P and T indicate bands for phosphorylated and total CREB. CREB phosphorylation levels in the nucleus accumbens (b, c) and caudate–putamen (e, f) at 10 min (b, e) and 20 min (c, f) following challenge injections using 2 × 2 factorial experimental design. Values are expressed as mean ± SEM (n = 6–8 per group) transformed to percentage of control values (repeated saline/saline challenge). †Indicates significantly lower levels compared with levels following saline challenge injections. ‡Indicates significantly higher levels compared with those following saline challenge injections or following cocaine challenge injections in rats repeatedly administered saline.

We also assessed CREB phosphorylation levels in rats that were previously assessed for locomotor activity using the 2 × 2 factorial design. CREB phosphorylation levels were analyzed separately for rats killed 10 and 20 min after challenge injections. The large number of samples prevented us from running all samples together on the same western blots. At 10 min, cocaine-induced CREB phosphorylation levels were significantly altered by repeated cocaine administration (two-way anova, effect of repeated administration, F1,24 = 5.11, p < 0.05; effect of challenge drug, F1,24 =5.72, p < 0.05; repeated administration × challenge drug, F1,24 = 4.63, p < 0.05) (Fig. 3b). Post hoc comparisons indicate that cocaine significantly increased levels by 33% following repeated administration of cocaine, but not following repeated administration of saline. At 20 min, cocaine-induced CREB phosphorylation levels were significantly altered by repeated cocaine administration (two-way anova, effect of repeated administration, F1,20 = 6.31, p < 0.05; effect of challenge drug, F1,20 = 7.23, p < 0.01; repeated administration × challenge drug, F1,20 = 5.13, p < 0.05) (Fig. 3c). Post hoc comparisons indicated that cocaine significantly increased levels three-fold following repeated administration of cocaine, but not following repeated administration of saline.

Cocaine-induced phosphorylation of CREB in the nucleus accumbens was also assessed by immunohistochemistry. Immunoreactivity was localized to the nucleus and evenly distributed within the nucleus accumbens (Fig. 4a). Repeated cocaine administration significantly altered the number of phosphorylated CREB-immunoreactive nuclei induced by cocaine (two-way anova, effect of repeated administration, F1,37 = 10.62, p < 0.01; effect of challenge drug, F1,37 =12.78, p < 0.01; repeated administration × challenge drug, F1,37 = 9.04, p < 0.01) (Fig. 4b). Post hoc comparisons indicated that cocaine significantly increased the number of phosphorylated CREB-immunoreactive nuclei two-fold following repeated administration of cocaine. Cocaine appeared to decrease the number of phosphorylated CREB-immunoreactive nuclei by 50% following repeated administration of saline, although this alteration was not statistically significant. Repeated administration of cocaine did not alter the number of phosphorylated CREB-immunoreactive nuclei induced by challenge injections with saline.

Figure 4.

Cocaine-dependent induction of phosphorylated CREB immunoreactivity in the nucleus accumbens is enhanced by repeated cocaine administration. (a) Photomicrographs (50 ×) of phosphorylated CREB immunoreactivity in the nucleus accumbens and caudate–putamen at + 1.8 mm from Bregma. Graphs represent the number of phosphorylated CREB-immunoreactive nuclei in the nucleus accumbens (b) and caudate-putamen (c). Values are expressed as mean ± SEM (n = 6–8 per group). †Indicates significantly higher levels compared with levels following saline challenge injections. ‡Indicates significantly higher levels of immunoreactive nuclei compared with levels following saline challenge injections or following cocaine challenge injections in rats repeatedly administered saline.

Caudate–putamen

In the time course experiment, repeated administration of cocaine did not significantly alter cocaine-induced CREB phosphorylation levels in the caudate–putamen (Fig. 3d). Challenge injections of cocaine did not significantly alter levels, although there was a trend for decreased levels at 10 and 60 min.

In the 2 × 2 factorial experiment, repeated cocaine administration did not alter cocaine-induced CREB phosphorylation levels 10 min after challenge injections (Fig. 3e). However, cocaine significantly decreased levels by 25% in rats repeatedly administered cocaine or saline (two-way anova, effect of challenge drug, F1,26 = 10.62, p < 0.01). At 20 min, phosphorylation levels were similar in all experimental groups (Fig. 3f).

Cocaine-induced phosphorylation of CREB in the caudate–putamen was also assessed using immunohistochemistry. Immunoreactivity was localized to the nucleus and evenly distributed throughout the caudate-putamen (Fig. 4a). Although cocaine challenge injections increased the number of phosphorylated CREB-immunoreactive nuclei (two-way anova, effect of challenge drug, F1,26 = 2.68, p < 0.05), repeated cocaine administration did not further enhance the number of immunoreactive nuclei (Fig. 4c). This finding contrasts with that of the western blot analysis, in which cocaine-induced CREB phosphorylation was not observed. This discrepancy is probably due to the different characteristics of each assay. The phospho-CREB antibody used in western blot and immunohistochemical experiments also recognizes the phosphorylated form of the CREB-related protein activating transcription factor-1 (ATF-1). Phosphorylated ATF-1 has a lower molecular weight than phosphorylated CREB, which allowed us to separate these two proteins in western blots. In contrast, immunohistochemical experiments cannot separate or otherwise distinguish between phosphorylated CREB and ATF-1. Thus, some of the increase in immunoreactive nuclei in immunohistochemical experiments might be due to cocaine-induced phosphorylation of ATF-1. For example, amphetamine has previously been shown to induce ATF-1 in striatal neurons (Genova and Hyman 1998). Another reason for the discrepancy is the inability of western blot experiments to distinguish CREB phosphorylation in activated neurons from that in non-activated neurons. In homogenates for western blot experiments, increased CREB phosphorylation levels in a minority of activated neurons can be diluted by unaltered or even decreased CREB phosphorylation levels in the majority of non-activated neurons. In contrast, phospho-CREB immunohistochemistry primarily detects increased phosphorylation in activated neurons without being diluted by decreased phosphorylation in surrounding neurons.

As cocaine-induced CREB phosphorylation was altered by repeated cocaine administration only in the nucleus accumbens and not in the caudate–putamen, we focused the remainder of our study on the nucleus accumbens.

Cocaine-induced ERK phosphorylation and activation in nucleus accumbens

ERK phosphorylation levels were calculated as the ratio of phosphorylated ERK to total ERK protein levels. The 42- and 44-kDa forms of phosphorylated ERK were quantified together. Total ERK protein levels were not significantly altered in any of our experiments (data not shown). In the time course experiment, repeated cocaine administration significantly altered cocaine-induced ERK phosphorylation levels in the nucleus accumbens (two-way anova, effect of repeated administration, F1,64 = 2.94, p < 0.05; effect of time, F4,64 = 4.98, p < 0.05; repeated administration × time, F4,66 = 2.92, p < 0.05)(Fig. 5a). Post hoc comparisons indicated that levels were higher in rats repeatedly administered cocaine than in rats repeatedly administered saline by 20% at 10 min and 100% at 20 min. ERK phosphorylation returned to basal levels within 40 min. Phosphorylation levels were not altered in rats repeatedly administered saline.

Figure 5.

Cocaine-induced ERK phosphorylation in the nucleus accumbens is enhanced following repeated cocaine administration. Time-dependent regulation of cocaine-induced ERK phosphorylation following repeated administration of cocaine or saline in the nucleus accumbens (a) and caudate-putamen (d). Values are expressed as mean ± SEM (n = 6–8 per group) transformed to percentage of control values (repeated saline/0-min control). †Indicates significantly higher levels at 10 and 20 min following repeated cocaine administration relative to levels following repeated saline administration. Blot insets show representative bands from western blots of cocaine-induced phosphorylated ERK following repeated saline administration (S) and repeated cocaine administration (C). P and T indicate bands for phosphorylated and total ERK. ERK phosphorylation levels in the nucleus accumbens (b, c) and caudate-putamen (e, f) at 10 min (b, e) and 20 min (c, f) following challenge injections using 2 × 2 factorial experimental design. Values are expressed as mean ± SEM (n = 6–8 per group) transformed to percentage of control values (repeated saline/saline challenge). †Indicates significantly higher levels following cocaine challenge injections compared with levels following saline challenge injections. ‡Indicates significantly higher levels compared with levels following saline challenge injections or following cocaine challenge injections in rats repeatedly administered saline.

ERK phosphorylation levels were also assessed 10 and 20 min following challenge injections using the 2 × 2 factorial experimental design. At 10 min, repeated cocaine administration did not significantly alter cocaine-induced ERK phosphorylation (Fig. 5b). However, challenge injections of cocaine increased ERK phosphorylation 25% above basal levels (two-way anova, effect of challenge drug, F1,24 = 6.05, p < 0.05). At 20 min, repeated administration with cocaine significantly altered cocaine-induced ERK phosphorylation levels (two-way anova, effect of repeated administration, F1,24 = 6.53, p < 0.01; effect of challenge drug, F1,24 = 4.73, p < 0.05; repeated administration ×challenge drug, F1,24 = 6.52, p < 0.01) (Fig. 5c). Post hoc comparisons indicated that cocaine increased ERK phosphorylation levels by 90% following repeated administration of cocaine, but not following repeated administration of saline.

ERK kinase activity levels were assessed in two additional sets of rats killed 10 and 20 min following challenge injections. An Elk-1 fusion protein was used as the substrate. Enzyme activity was indicated by the levels of Elk-1 phosphorylation in subsequent western blot assays. At 10 min, repeated cocaine administration did not significantly alter cocaine-induced ERK kinase activity levels (Fig. 6a). However, challenge injections of cocaine significantly increased levels by 80% in rats repeatedly administered cocaine or saline (two-way anova, effect of challenge drug, F1,8 = 16.22, p < 0.01). At 20 min, repeated cocaine administration significantly enhanced cocaine-induced ERK kinase activity (two-way anova, effect of repeated administration, F1,11 = 27.02, p < 0.01; effect of challenge drug, F1,11 = 52.63, p < 0.001; repeated administration × challenge drug, F1,11 = 5.70, p < 0.05) (Fig. 6b). Post hoc comparisons indicated that cocaine increased activity levels almost 100% following repeated administration of saline and 200% following repeated administration of cocaine.

Figure 6.

Cocaine-induced ERK activity is enhanced in the nucleus accumbens following repeated cocaine administration. ERK activity levels in the nucleus accumbens at 10 min (a) and 20 min (b) following challenge injections using 2 × 2 factorial experimental design. Values are expressed as mean ± SEM (n = 6–8 per group) transformed to percentage of control values (repeated saline/saline challenge). †Indicates significantly higher levels compared with levels following saline challenge injections. ‡Indicates significantly higher levels compared with levels following saline challenge injections or following cocaine challenge injections in rats repeatedly administered saline.

Cocaine-induced phosphorylation of CaMKs II and IV and and GluR1 S845

CaMKs II and IV and PKA can phosphorylate Ser133 in CREB. As cocaine-induced CREB phosphorylation in the nucleus accumbens was enhanced 10 and 20 min following challenge injections, we assessed endogenous activity levels of CaMKs II and IV and PKA at these two time points. In all cases, phosphorylation levels were calculated as the ratio of phosphorylated protein to total protein levels. Total protein levels for CaMKs II and IV, and GluR1 were not significantly altered in any of our experiments (data not shown). CaMKII activity levels were assessed by measuring phosphorylation of Thr286 in CaMKII, which is autophosphorylated when the enzyme is activated. At 10 min, repeated cocaine administration did not alter cocaine-induced CaMKII phosphorylation levels (Fig. 7a). However, cocaine challenge injections increased phosphorylation levels by 50% in rats repeatedly administered cocaine or saline (two-way anova, effect of challenge drug, F1,23 = 11.02, p < 0.01). At 20 min, CaMKII phosphorylation levels were similar in all groups (Fig. 7b).

Figure 7.

CaMKII, CaMKIV and GluR1 Ser845 phosphorylation levels in the nucleus accumbens. Levels of CaMKII phosphorylation in the nucleus accumbens at 10 min (a) and 20 min (b) following challenge injections using 2 × 2 factorial experimental design. Values are expressed as mean ± SEM (n = 6–8 per group) transformed to percentage of control values (repeated saline/saline challenge). †Indicates significantly higher levels following cocaine challenge injections compared with levels following saline challenge injections. Levels of CaMKIV phosphorylation in the nucleus accumbens at 10 min (c) and 20 min (d) following challenge injections using 2 × 2 factorial experimental design. Values are expressed as mean ± SEM (n = 6–7 per group) transformed to percentage of control values (repeated saline/saline challenge). †Indicates significantly lower levels compared with levels following saline challenge injections in rats repeatedly administered saline. ‡Indicates significantly higher levels following cocaine challenge injections compared to levels following cocaine challenge injections in rats repeatedly administered saline. Levels of GluR1 Ser845 phosphorylation in the nucleus accumbens at 10 min (e) and 20 min (f) following challenge injections using 2 × 2 factorial experimental design. Values are expressed as mean ± SEM (n = 6–8 per group) transformed to percentage of control values (repeated saline/saline challenge). †Indicates significantly higher levels following cocaine challenge injections compared with levels following saline challenge injections.

CaMKIV activity levels were assessed by measuring phosphorylation of Thr196 in CaMKIV, which is phosphorylated by CaMK kinase to activate the enzyme. At 10 min, repeated cocaine administration significantly altered cocaine-induced CaMKIV phosphorylation levels (two-way anova, effect of repeated administration, F1,21 = 3.50, p = 0.08; effect of challenge drug, F1,21 = 6.09, p < 0.05; repeated administration × challenge drug, F1,21 = 10.5, p < 0.005) (Fig. 7c). Post hoc comparisons indicated that phosphorylation levels following cocaine challenge injections were 61% higher after repeated cocaine administration than after repeated saline administration. However, it should be noted that this relative increase was due to cocaine significantly decreasing CaMKIV phosphorylation levels after repeated saline administration but having no significant effect after repeated cocaine administration. At 20 min, repeated cocaine administration did not alter cocaine-induced CaMKIV phosphorylation levels (Fig. 7d). However, cocaine challenge injections still significantly decreased phosphorylation levels (two-way anova, effect of challenge drug, F1,10 = 6.67, p < 0.05).

Endogenous PKA activity levels were assessed by measuring PKA-specific phosphorylation of Ser845 in the GluR1 subunit of the α-amino-3-hydroxy-5-methylisoxazole-4-propionate receptor (Roche et al. 1996; Mammen et al. 1997; Chao et al. 2002). At 10 min, repeated cocaine administration did not alter levels of cocaine-induced phosphorylation at Ser845 (Fig. 7e). However, cocaine challenge injections increased phosphorylation levels by 25% in rats repeatedly administered cocaine or saline (two-way anova, effect of challenge drug, F1,27 = 17.88, p < 0.01). At 20 min, repeated cocaine administration did not alter levels of cocaine-induced phosphorylation at Ser845 (Fig. 7f). However, cocaine challenge injections increased phosphorylation levels by 30% in rats repeatedly administered cocaine and saline (two-way anova, effect of challenge drug, F1,28 = 25.32, p < 0.01).

U0126 infusions into the nucleus accumbens attenuate CREB phosphorylation

Intracranial experiment 1: U0126 infusions and repeated drug administration

Based on our data, we hypothesized that repeated cocaine administration activates cocaine-induced CREB phosphorylation by enhancing ERK phosphorylation and activation of ERK kinase activity. To test our hypothesis, we unilaterally infused U0126, a MEK inhibitor, into the nucleus accumbens to block cocaine-induced ERK and CREB phosphorylation after repeated administration of either cocaine or saline. We used a 2 × 2 factorial experimental design in which levels of phosphorylation were compared between the U0126-infused side and the vehicle-infused side in nucleus accumbens tissue. Vehicle infusions into contralateral hemispheres controlled for individual differences between rats and handling experiences. Tissue was obtained 20 min after cocaine challenge injections in rats repeatedly administered cocaine or saline.

U0126 altered cocaine-induced ERK phosphorylation by approximately 60% in rats repeatedly administered with cocaine or saline (two-way anova, effect of repeated administration, F1,24 = 4.35, p < 0.05; effect of U0126, F1,24 = 9.94, p < 0.01; repeated administration × U0126, F1,24 = 6.51, p = 0.05) (Fig. 8a). Post hoc comparisons indicated that cocaine-induced ERK phosphorylation was 75% lower in the U0126-infused hemisphere in rats that received repeated administrations of cocaine. Cocaine-induced ERK phosphorylation was not significantly different between U0126-infused and vehicle-infused hemispheres in rats that received repeated administrations of saline. These results confirm that cocaine-dependent CREB phosphorylation in the nucleus accumbens is mediated by ERK activity. Results from the vehicle-infused hemispheres replicated our previous finding that cocaine-induced ERK phosphorylation is enhanced in the nucleus accumbens following repeated cocaine administration. Cocaine-induced ERK phosphorylation levels in the vehicle-infused hemispheres were approximately 20% higher in rats repeatedly administered cocaine than in rats repeatedly administered saline. It should be noted that U0126 did not alter phosphorylation of the PKA-specific site Ser845 in GluR1 (data not shown); this finding demonstrates the specific inhibitory action of U0126 in our experiments.

Figure 8.

Intra-accumbens infusions of U0126 attenuate cocaine-induced phosphorylation of (a) ERK and (b) CREB in the nucleus accumbens following repeated administration of cocaine or saline. Values are expressed as mean ± SEM (n = 6 per group) transformed to percentage of control values (repeated saline/vehicle-infused hemisphere) for each hemisphere. †Indicates significantly lower levels compared with levels in the vehicle-infused hemisphere. ‡Indicates significantly higher levels in the vehicle-infused hemisphere of rats following repeated cocaine administration relative to levels in the vehicle-infused hemisphere of rats following repeated saline administration. (c) Intra-accumbens infusions of U0126 attenuate cocaine-induced, but not saline-induced, CREB phosphorylation levels in the nucleus accumbens after repeated cocaine administration to all rats. Values are expressed as mean ± SEM (n = 4–6 per group) transformed to percentage of control values (saline challenge/vehicle-infused hemisphere) for each hemisphere. ‡Indicates significantly lower levels compared with levels in the vehicle-infused hemisphere.

U0126 altered cocaine-induced CREB phosphorylation in the nucleus accumbens of rats repeatedly administered with cocaine or saline (two-way anova, effect of repeated administration, F1,15 = 4.96, p < 0.05; effect of U0126, F1,15 = 24.04, p < 0.01; repeated administration × U0126, F1,15 = 5.71, p < 0.05) (Fig. 8b)Post hoc comparisons indicated that cocaine-induced CREB phosphorylation was 66% lower in the U0126-infused hemisphere in rats that received repeated cocaine administration. Cocaine-induced CREB phosphorylation was approximately 30% lower in the U0126-infused hemisphere in rats that received repeated saline administration. Again, results from the vehicle-infused hemispheres replicated our previous finding that cocaine induces CREB phosphorylation in the nucleus accumbens following repeated cocaine administration. Cocaine-induced CREB phosphorylation levels in the vehicle-infused hemispheres were approximately 40% higher in rats repeatedly administered cocaine than in rats repeatedly administered saline.

U0126 decreased cocaine-induced phosphorylation of ERK and CREB in the nucleus accumbens. Cross-correlation analysis of phosphorylation levels from the vehicle and U1026 hemispheres of rats repeatedly administered cocaine revealed a strong relationship between cocaine-induced CREB phosphorylation and ERK phosphorylation (r2 =0.88; p < 0.001).

Intracranial experiment 2: U0126 infusions and challenge injections

U0126 or vehicle was infused into the nucleus accumbens 30 min before cocaine or saline challenge injections to rats that all received repeated administrations of cocaine. This experiment was used to confirm that U0126 blocked cocaine-dependent effects on CREB phosphorylation rather than altering basal phosphorylation levels, which was not controlled for in intracranial experiment 1. U0126 infusions significantly reduced cocaine-dependent CREB phosphorylation in the nucleus accumbens (two-way anova, challenge drug × U0126, F1,17 = 4.70, p < 0.05) (Fig. 8c). Post hoc comparisons indicated that CREB phosphorylation levels in cocaine-challenged rats were 38% lower in the U0126-infused hemisphere whereas ERK phosphorylation levels were 23% lower. CREB phosphorylation levels in saline-challenged rats were not significantly altered by U0126 even though ERK phosphorylation levels were significantly reduced by 30%. U0126 infusions did not have a significant effect on GluR1 phosphorylation levels in the same samples.

Intracranial experiment 3: Rp-cAMPs infusions and challenge injections

Rp-cAMPs or vehicle was infused into the nucleus accumbens of rats 30 min before cocaine challenge injections to rats that received repeated administrations of cocaine. Rp-cAMPs infusions did not alter CREB phosphorylation levels in the nucleus accumbens (Fig. 9a) even though Rp-cAMPs significantly decreased GluR1 phosphorylation levels by 45% (Fig. 9b).

Figure 9.

Intra-accumbens infusions of Rp-cAMPs did not alter CREB phosphorylation (a) but did attenuate GluR1 phosphorylation (b) in the nucleus accumbens of rats following cocaine challenge injections after repeated cocaine administration. Grey bars indicate cocaine-induced levels of phosphorylation in the vehicle-infused hemispheres and black bars indicate levels of phosphorylation in the U0126-infused hemispheres. Values are expressed as mean ± SEM (n = 6 per group) transformed to percentage of control values (vehicle-infused hemisphere) for each hemisphere. †Indicates significantly lower levels compared with levels in the vehicle-infused hemisphere.

Discussion

Cocaine induced CREB phosphorylation in the nucleus accumbens of rats sensitized by repeated administration of cocaine, but not in non-sensitized rats that received repeated saline administration. It appears that repeated cocaine administration enables a mechanism that allows cocaine to induce CREB phosphorylation. At the time points we assessed, cocaine-induced activation of ERK, but not PKA or CaMKs II and IV, was also enhanced in these sensitized rats. Based on these findings, we hypothesized that enhanced ERK activation enables and mediates cocaine-induced CREB phosphorylation in the nucleus accumbens of cocaine-sensitized rats. This was confirmed when intra-accumbens infusions of the MEK inhibitor U0126, but not the PKA inhibitor Rp-cAMPs, blocked cocaine-induced CREB phosphorylation in cocaine-sensitized rats.

We found that an intermediate dose (15 mg/kg) of cocaine did not induce CREB phosphorylation in the nucleus accumbens 7 days after repeated saline administration. In contrast, Simpson et al. (1995) found that a high dose (5 mg/kg) of acute amphetamine increased phospho-CREB immunoreactivity in the core subregion of rat nucleus accumbens. This is probably due to the stronger neurochemical effects of high doses of amphetamine compared with the intermediate dose of cocaine used in our study (Carboni et al. 1989; Vanderschuren and Kalivas 2000). Simpson et al. (1995) also found an attenuation of amphetamine-induced phospho-CREB immunoreactivity 1 day after repeated amphetamine administration. In contrast, we found that cocaine-induced CREB phosphorylation levels were higher in the nucleus accumbens 7 days after repeated cocaine administration. The different withdrawal times in these studies are likely to explain the different effects on drug-induced CREB phosphorylation. Short withdrawal times of 1–2 days are often associated with tolerant intracellular signaling within the striatum (Hope et al. 1992; Persico et al. 1993; Steiner and Gerfen 1993) that is absent at later withdrawal times (Nye et al. 1995; Persico et al. 1995). For the caudate–putamen, we found that an intermediate dose (15 mg/kg) of cocaine did not induce CREB phosphorylation 7 days after repeated administration of saline or cocaine. In contrast, high doses of amphetamine (Cole et al. 1995; Simpson et al. 1995; Turgeon et al. 1997) and cocaine (Kano et al. 1995) have been shown to induce CREB phosphorylation in the striatum. The different results can most probably be explained by the different doses of drugs used. Another explanation may be drug administration in the rat's home cage in these latter studies versus administration outside the home cage in our study; however, a direct comparison of the effects of drug environment remains to be done.

After repeated cocaine administration to rats outside their home cage, we found that cocaine-induced ERK phosphorylation and activity were increased in the nucleus accumbens at 10 min and enhanced at 20 min following cocaine challenge injections. This is consistent with the results of Valjent et al. (2000), who found that a single injection of cocaine increased phospho-ERK immunoreactivity at 10 min, but not 20 min, whereas a challenge injection of cocaine 1 day after repeated cocaine administration in the home cage increased immunoreactivity at both 10 and 20 min. Although the comparison with our study is indirect, results of Valjent et al. (2000) suggest that enhanced cocaine-induced ERK phosphorylation is not specific to repeated drug administration outside the rat's home cage.

As ERK was the only kinase whose activation was enhanced following repeated cocaine administration, we hypothesized that cocaine-induced CREB phosphorylation in the nucleus accumbens is enabled by enhanced activation of ERK. Phosphorylated ERK translocates to the nucleus to phosphorylate CREB via ribosomal protein S6 kinase (Rsk2) (Chen et al. 1992; Blenis 1993; Impey et al. 1998). In the present study, intra-accumbens infusions of the MEK inhibitor U0126 blocked the increase in cocaine-induced ERK and CREB phosphorylation levels in cocaine-sensitized rats. Thus enhanced cocaine-induced ERK activation enables and mediates cocaine-induced CREB phosphorylation in the nucleus accumbens of cocaine-sensitized rats.

In the present study, we found that infusions of Rp-cAMPs did not attenuate cocaine-induced CREB phosphorylation in nucleus accumbens of cocaine-sensitized rats. This indicates that PKA is not responsible for activating cocaine-induced CREB phosphorylation in the nucleus accumbens of conscious, behaving rats. These data may appear to contradict our previous finding that infusions of Rp-cAMPs into the nucleus accumbens blocked amphetamine-induced CREB phosphorylation in the nucleus accumbens (Self et al. 1998). However, this result was obtained using anesthetized rats, and regulation of intracellular mechanisms associated with CREB phosphorylation in anesthetized rats is likely to be very different from that in conscious, behaving rats. Neural activity and ERK phosphorylation levels in the striatum are increased by electrical activation of excitatory glutamatergic afferents from the cortex to the striatum (Sgambato et al. 1998). Transection of these corticostriatal afferents in conscious, behaving rats blocks amphetamine-induced ERK phosphorylation in striatum (Ferguson and Robinson 2004). When rats are anesthetized, excitatory glutamatergic neurotransmission from these corticostriatal afferents (West 1998; Kreuter et al. 2004) and cocaine-induced neural activity in the striatum are attenuated (Ryabinin et al. 2000; Kreuter et al. 2004). Therefore, cocaine-induced ERK activity and ERK-dependent CREB phosphorylation would also be reduced in anesthetized rats. In the absence of significant ERK activation, the influence of basal levels of PKA-dependent CREB phosphorylation in anesthetized rats would appear greater than it is in conscious, behaving rats.

We have also shown that PKA activity levels increase in the nucleus accumbens after repeated cocaine administration (Terwilliger et al. 1991; Hope et al. 2005) and remain increased at 7 days of withdrawal (Hope et al. 2005), the time point at which we examined levels of endogenous PKA activity and CREB phosphorylation in the present study. However, the in vitro PKA assays used in Terwilliger et al. (1991) and Hope et al. (2005) assessed maximum levels of PKA activity capable of being activated in nucleus accumbens. These assays did not assess levels of endogenous functional activity of PKA at the time when brain tissue was obtained. To address this issue in the present study, we used phosphorylation of GluR1 Ser845 as an indicator of endogenous functional activity of PKA (Roche et al. 1996; Mammen et al. 1997; Chao et al. 2002). We found that cocaine clearly induced PKA activity at 10 and 20 min following challenge injections, but this activity was not further enhanced in nucleus accumbens of cocaine-sensitized rats.

CaMKs II and IV can phosphorylate CREB at Ser133 (Schulman 1993; Matthews et al. 1994). In our study, cocaine increased CaMKII phosphorylation levels 10 min following challenge injections, but these levels were not further enhanced in the nucleus accumbens by repeated cocaine administration. In cultured neurons and hippocampal slice preparations, CaMKIV-dependent CREB phosphorylation peaks and declines within 10 min after stimulation, whereas more sustained CREB phosphorylation is mediated by slower activated ERK activity (Matthews et al. 1994; Finkbeiner et al. 1997; Deisseroth et al. 1998; Kasahara et al. 2001; Wu et al. 2001). CaMKIV could be performing a similar role shortly after cocaine administration. In our study, cocaine-induced CaMKIV phosphorylation levels were significantly higher following repeated cocaine administration. However, even in the nucleus accumbens of rats that were repeatedly administered cocaine, cocaine challenge injections did not increase phosphorylation levels in the nucleus accumbens above those induced by saline challenge injections. It might therefore appear that CaMKs II and IV cannot contribute to activation of cocaine-induced CREB phosphorylation in nucleus accumbens of cocaine-sensitized rats. However, phosphorylation levels of CaMKs II and IV may have peaked and declined within 10 min of the cocaine challenge injection, similar to observations in hippocampal slices. If CaMK activity levels are enhanced within this 10-min time interval, CREB phosphorylation might also be augmented in this period and persist even after the CaMKs have been dephosphorylated and inactivated.

Increased activation of ERK kinase and CREB phosphorylation can also explain enhanced induction of Fos in the nucleus accumbens of behaviorally sensitized rats (Crombag et al. 2002; Todtenkopf et al. 2002). C-fos transcription is activated by ERK-dependent phosphorylation of CREB and Elk-1 on the c-fos promoter (Cavigelli et al. 1995; Davis et al. 2000; Mayr and Montminy 2001; Kornhauser et al. 2002). Increased cocaine-induced CREB phosphorylation in the nucleus accumbens, but not in caudate–putamen, can explain the selective enhancement of cocaine-induced Fos expression in the nucleus accumbens. We were unable to measure Elk-1 phosphorylation directly, but we found enhanced cocaine-induced activation of ERK-dependent phosphorylation of the Elk-1 substrate in the nucleus accumbens of behaviorally sensitized rats. Enhanced Fos induction in the nucleus accumbens is therefore most likely the result of enhanced ERK-dependent activation of CREB and Elk-1 phosphorylation.

Enhanced cocaine-induced Fos expression in the nucleus accumbens 7 and 14 days after repeated cocaine administration (Crombag et al. 2002; Todtenkopf et al. 2002) suggests increased activation of neurons in this brain region. Enhanced neural activation is likely to result in enhanced calcium influx that can then activate ERK through a Ras-dependent pathway and phosphorylate CREB and Elk-1 to increase Fos induction (Heist and Schulman 1998; Deisseroth and Tsien 2002; Lisman et al. 2002; Sweatt 2004; Thomas and Huganir 2004). It is likely that common mechanisms for enhanced ERK activation and enhanced neural activity will be found upstream of ERK activation. Because neural activity in the nucleus accumbens is critical for cocaine-induced locomotor activity (Kelly and Iversen 1976; Delfs and Kelley 1990) as well as reward and reinforcement (Wise and Bozarth 1987; Koob 1992; Wise 2004), identification of the upstream mechanisms that produce enhanced ERK activation in the nucleus accumbens should also help to reveal how repeated drug administration alters these drug-induced behaviors.

Acknowledgements

We would like to thank Jeremy Gilbert and Arlene Pak for technical assistance. We also thank Emily Wentzell for editorial assistance.

Ancillary