|Home | About | Journals | Submit | Contact Us | Français|
The interactions between dopamine and glutamate systems play an essential role in normal brain functions and neuropsychiatric disorders. The mechanism of NMDA receptor regulation through high concentrations of dopamine, however, remains unclear. Here we show the signaling pathways involved in hyperdopaminergic regulation of NMDA receptor functions in the prefrontal cortex (PFC) by incubating cortical slices with high concentration of dopamine or administering dopamine reuptake inhibitor GBR12909 in vivo. We found that, under both conditions, the synaptic NMDA receptor-mediated currents were significantly attenuated by excessive dopamine stimulation through activation of D2 receptors. Furthermore, high-dose of dopamine failed to affect NMDA receptor-mediated currents after blockade of NR2B subunits but triggered a dynamin-dependent endocytosis of NMDA receptors. The high-dose dopamine/D2 receptor-mediated suppression of NMDA receptors was involved in the increase of GSK-3β activity, which in turn phosphorylates β-catenin and disrupts β-catenin-NR2B interaction, but was neither dependent on Gq11 nor PLC. Moreover, the hyperdopamine induced by GBR12909 significantly decreased the expression of both surface and intracellular NR2B proteins, as well as NR2B mRNA levels, suggesting an inhibition of protein synthesis. These effects were, however, completely reversed by administration of either GSK-3β inhibitor or D2 receptor antagonist. These results therefore suggest that GSK-3β is required for the hyperdopamine/D2 receptor-mediated inhibition of NMDA receptors in the prefrontal neurons and these actions may underlie D2 receptor-mediated psychostimulant effects and hyperdopamine-dependent behaviors in the brain.
The prefrontal cortex (PFC) is a brain region highly associated with cognition and psychiatric disorders (Harrison and Weinberger, 2005; Lewis and Gonzalez-Burgos, 2008). Dopamine (DA) as a major neurotransmitter in the PFC has long been implicated in schizophrenia (Carlsson, 1978). Excess of DA or an elevated sensitivity to DA stimulation profoundly affects the synaptic transmission in the brain (Greengard, 2001; Seamans and Yang, 2004; Seeman, 2006). Recent studies suggested that many psychiatric disorders are strongly associated with abnormalities in both DA and glutamate systems. Indeed, besides DA, evidence supports the hypothesis that schizophrenia might be related to NMDA receptor hypofunction (Jentsch and Roth, 1999; Krystal et al., 2002; Moghaddam and Jackson, 2003). In the past decade, extensive studies have been focused on the interactions between DA and NMDA receptors (Goldman-Rakic et al., 2004; Yang and Chen, 2005; Cepeda and Levine, 2006). It is widely believed that optimal DA preferentially acts on D1 receptors to promote NMDA receptor function (Seamans et al., 2001; Flores-Hernandez et al., 2002), possibly through receptor phosphorylation or trafficking (Dunah and Standaert, 2001; Dunah et al., 2004; Hallett et al., 2006; Gao and Wolf, 2008; Tong and Gibb, 2008) (Li et al, unpublished observation). In contrast, high level of DA or activation of D2 receptors may attenuate NMDA receptor-mediated transmission (Zheng et al., 1999; Wang et al., 2003; Beazely et al., 2006; Liu et al., 2006; Jiao et al., 2007; Martina and Bergeron, 2008). The fundamental problem is that most of the studies focused on the D1-NMDA receptor interaction and interpreted it as a hypothetical inverted “U” curve for D1 action (Goldman-Rakic et al., 2000; Seamans and Yang, 2004). The mechanisms involved in the D2 receptor-dependent effects on NMDA receptor functions under conditions of hyperdopaminergic state are, however, not well investigated.
Recent studies have provided interesting insights into the involvement of glycogen synthase kinase-3β (GSK-3β) signaling in schizophrenia (Kozlovsky et al., 2000; Emamian et al., 2004; Koros and Dorner-Ciossek, 2007; Lovestone et al., 2007). The GSK-3β signaling is also particularly important in D2 receptor-mediated psychostimulant effects and DA dependent behaviors (Beaulieu et al., 2005; Bibb, 2005; Beaulieu et al., 2007a; Beaulieu et al., 2007b). In addition, antipsychotic agents can decrease the activity of GSK-3β by increasing GSK-3β Ser9 (Emamian et al., 2004; Li et al., 2007). GSK-3 is a serine/threonine protein kinase known for phosphorylating and thus inactivating glycogen synthase. GSK-3 contains two isoforms, GSK-3α and GSK-3β, but only GSK-3β is highly enriched in the brain (Woodgett, 2001). Because GSK-3β inhibitor is also involved in the regulation of NMDA receptor-dependent synaptic plasticity (Peineau et al., 2007; Zhu et al., 2007; Peineau et al., 2008), we hypothesized that hyperdopamine/D2 receptor-induced inhibition of NMDA receptors may occur through GSK-3β-mediated NMDA receptor trafficking, which would result in NMDA receptor hypofunction. We tested this hypothesis using various approaches and demonstrated that elevated dopamine via D2 receptors triggered the internalization of NR2B subunits and this process required the activation of GSK-3β.
Fifty Sprague-Dawley rats at ages of postnatal day (PD) 12 to 30 were used for this study. The animals were treated under the National Institute of Health (NIH) guidelines and the experimental protocol was approved by the Institutional Animal Care and Use Committee at Drexel University College of Medicine. The detailed procedure can be found in our previous studies (Gao et al., 2001; Gao and Goldman-Rakic, 2003; Gao, 2007). The animals were deeply anesthetized with Euthasol (0.2 ml/kg, Virbac AH, Texas, USA), and the brains were immediately removed and placed in ice-cold (< 4°C) sucrose (in mM, 2.5 KCl, 1.25 NaH2PO4, 26 NaHCO3, 0.5 CaCl2, 7.0 MgSO4, 213 sucrose; pH 7.4) solution buffered with 95% O2 and 5% CO2. The blocks of neocortex containing medial PFC (PrL, Paxinos and Watson, The Rat Brain in Stereotaxic Coordinates. 2005) were trimmed and were sectioned using Leica VT1000s Vibratome (Leica Microsystems USA, Bannockburn, IL, USA). The horizontal brain slices at thickness of 300 μm were incubated in oxygenated Ringer’s solution at 35°C for 1 hour. Ringer’s solution contains the following ingredients (in mM): 128 NaCl, 2.5 KCl, 1.25 NaH2PO4, 2 CaCl2, 1 MgSO4, 26 NaHCO3, and 10 dextrose, pH 7.4. The slices were then kept at room temperature until being transferred to a submerged chamber for recording. Whole-cell patch clamp recordings were conducted in the medial PFC through an upright Zeiss Axioskop 2 microscope (Carl Zeiss, Inc., Germany) equipped with optics of infrared-differential interference contrast and a digital video camera system. The recordings were conducted at ~ 35°C. Resistance of the recording pipette (1.2 mm borosilicate glass, Warner Instrument, Hamden, CT) was 7–9 MΩ. For voltage clamp, a Cs+-solution containing (in mM) Cs-gluconate 120, lidocaine 5 (QX- 314), 6 CsCl2, 1 ATP-Mg, 0.2 Na2GTP, and 10 Hepes at pH 7.3 (adjusted with CsOH) was used to block sodium and potassium channels. The NMDA receptor-mediated excitatory postsynaptic currents (EPSCs) were recorded at +60 mV by stimulating layer 2/3 with a bipolar electrode placed ~ 200–500 μm from the recording neurons (single pulse, 0.1 ms, 10–100 μA, 0.1 Hz) in the presence of picrotoxin (50 μM) and CNQX (20 μM). To record the spontaneous NMDA receptor-mediated EPSCs (sEPSC), the neurons were held at a membrane potential of +60 mV with bath perfusion of picrotoxin (50 μM) and NBQX (10 μM). All chemicals were purchased from Sigma-Aldrich unless specified (Sigma-Aldrich Inc., St. Louis, MO). The electric signals were amplified and filtered at 2 kHz in voltage clamp mode with MultiClamp 700B (Molecular Devices, Union City, CA) and acquired at sampling intervals of 20-50 μs through a DigiData 1322A (data acquisition system) and pCLAMP 9.2 software (Molecular Devices, Union City, CA). The series resistances were compensated and were constantly monitored by injection of a hyperpolarization current pulse (−0.1 nA) at duration of 150 ms.
Only cells exhibiting stable membrane potentials without injection of holding current and producing stable EPSCs without rundown for at least 5 minutes were used for further analysis of drug effects. The amplitudes of the evoked EPSCs were measured by averaging 30 traces from the onset to the peak of EPSC with Clampfit 9.2 software (Molecular Devices, Union City, CA). The sEPSCs recorded in voltage-clamp mode were analyzed with Clampfit 9.2 (Molecular Devices, Union City, CA). A typical sEPSC was selected to create a sample template for event detection within a data period. The frequency (event number) and amplitude of individual events were examined with Clampfit. The data were analyzed by Student T-test and were presented as mean ± standard error.
Bis (sulfosuccinimidyl) suberate(BS3) crosslinking was performed as described previously (Grosshans et al., 2001, 2002; Conrad et al., 2008). BS3 is a membrane-impermeable agent, which selectively crosslinks cell-surface proteins to form high-molecular-mass aggregates. Because intracellular proteins are not modified, they retain normal molecular mass. This enables surface and intracellular pools of a particular protein to be distinguished by SDS–PAGE and Western blotting. Brain tissues containing mPFC were quickly sectioned as 400-μm slices with a Vibratome, and slices were incubated with BS3 (1 mg/ml; Pierce Biotechnology, Rockford, IL, USA) in aerated ACSF (95% O2 and 5% CO2) at 4°C for 40 minutes with gentle agitation. The slices were then washed three times with ice-cold ACSF containing 20 mM Tris (pH 7.6) to quench the remaining BS3 and the surface expression was determined by Western blot analysis.
Tissues containing PFC were microdissected and then homogenized in ice-cold lysis buffer (50 mM Tris-HCl pH 8.0, 150 mM NaCl, 1% NP-40, protease inhibitor cocktail), centrifuged at 13,000 × g for 10 minutes at 4°C. Supernatant fractions (approximately 500 μg proteins) were incubated overnight with 2.5 μg of anti-β-catenin (Cell Signaling Technology, Danvers, MA). The immuno-complexes were isolated by addition of 20 μl of protein G– Sepharose beads (GE Healthcare Bio-Science AB, Piscataway, NJ, USA), followed by incubation for 3–4 h at 4°C. The immunoprecipitates were then washed four times with lysis buffer, resuspended in laemmli sample buffer, and boiled for 10 minutes. After they were centrifuged at 10,000 × g for 5 minutes, supernatant was collected. The immunoprecipitated proteins were analyzed by Western blot analysis with antibodies against NR2B (1:2000, Millipore-Chemicon, Millipore Corp., Billerica, MA), β-catenin, and β-tubulin.
PFC tissues were homogenized in a lysis buffer (20 mM Tris-HCl with pH 7.4, 200 mM NaCl, 1 mM Na3VO4, 10 mM NaF, and protease inhibitor cocktail), and centrifuged at 13,000 × g for 2 minutes. Supernatant fraction was aliquoted and stored at −80°C. Equal amounts of proteins were subjected to 7.5% SDS-polyacrylamide gel electrophoresis. Proteins were transferred to a nitrocellulose membrane (Bio-Rad Laboratories, Hercules, CA, USA) and blocked with 5% non-fat milk in TBS-T (0.05% Tween-20 in 1X Tris-buffered saline). The blots were incubated with anti-NR2B, anti-NR2A (Millipore-Chemicon, Millipore Corp., Billerica, MA), anti-GSK-3β, anti-phosphor-GSK-3β Ser9, anti-β-catenin, anti-phospho-β-catenin (Ser33/37/Thr41) (Cell Signaling Technology, Danvers, MA), or anti-phosphor-GSK-3β Tyr216 (Ambgood, Richmond, BC, Canada) at a dilution of 1:2,000 in TBS-T with 5% milk. Anti-NR2B-pS1480 antibody (gifted from Dr. Richard L. Huganir) was used at 1:200 (Chung et al., 2004). After several washes with TBS-T, the blots were then incubated in horseradish peroxidase (HRP)-conjugated goat anti-mouse IgG (Jackson ImmunoResearch Laboratories, Inc., West Grove, PA, USA) at 1:2,000 for 2 hours. The immunopositive protein bands were detected with ECL Western Blotting System (Amersham Bioscience, Piscataway, NJ). After the exposure of membranes to Kodak Biomax film(Eastman Kodak, Rochester, NY), the band densities were measured with Image J software (NIH). Final data were normalized to the levels of β-actin or total proteins (for phosphorylation). To minimize the inter-blot variability, each sample was analyzed 4 times. The mean value for each sample was calculated from all the replicates in different animals and the results were presented as mean ± standard error. Significance was determined with Student T-test or ANOVA.
Total RNA was extracted from the homogenized lysis mixture directly using an RNAqueous Micro Kit and eventually was dissolved in 40 μl elution buffer for the PFC tissue sample. RNA concentration was measured at wavelengths of 260 and 280 nm using the NanoDrop ND-1000 Spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA). Only the RNA samples with 260/280 ratios of 1.8 to 2.0 were used for further analyses. RNA extracted from the PFC tissue was directly diluted into RNase-free water at a concentration of 20 ng/μl. The Qiagen one-step RT-PCR kit (Qiagen, Germantown, MD, USA) was used to perform RT-PCR with glyceraldehyde-3-phosphate dehydrogenase (GAPDH) as housekeeping gene. The primers used are: GAPDH (NM_017008) forward primer ccatcccagaccccataac, backward primer gcagcgaactttattgatgg, product 78 bp; GRIN2B (NR2B, NM_012574) forward primer tgagtgagggaagagagagagg, backward primer atggaaacaggaatggtgga, product 249 bp. The detailed RT-PCR set-up was as follows: 1.0 μl RNA, 2.0 μl NR2B forward primer (4 μM), 2.0 μl NR2B backward primer (4 μM), 2.0 μl GAPDH forward primer (4 μM), 2.0 μl GAPDH backward primer (4 μM), 5.0 μl Q buffer, 5.0 μl 5X buffer, 1.0 μl dNTP, 1.0 μl enzyme, 4.0 μl RNase-free water. The PCR protocol was: 1) reverse transcription, 50°C for 30 min; 2) initial PCR activation, 95 °C for 15 min; 3) three-step cycling (30 cycles), 94 °C for 1 min, 55 °C for 1 min, 72 °C for 1 min; 4) Final extension, 72 °C for 10 min. PCR products were evaluated by gel electrophoresis and the mRNA expression levels for NR2B were normalized to the individual GAPDH level first, then to the control level.
The synaptically evoked NMDA receptor-mediated EPSCs were recorded from layer 5 pyramidal neurons in the medial PFC by stimulating layer 2/3, with the membrane potential held at +60 mV and with the presence of AMPA receptor antagonist CNQX (20 μM) and GABAA receptor antagonist picrotoxin (50 μM). The NMDA receptor-mediated EPSCs were confirmed in some cases and were completely abolished by bath application of the NMDA receptor antagonist D-APV (50 μM, data not shown). To examine the effects of high level of DA in NMDA receptors, we first bath applied DA at different concentrations from 1 μM to 200 μM and compared the DA actions in NMDA receptor-mediated EPSCs. We found that DA exhibited bidirectional dose-dependent effects on NMDA receptor-mediated EPSCs in the layer 5 pyramidal neurons. As shown in Figure 1, low concentrations of DA (1 μM) significantly increased the NMDA receptor-mediated EPSCs by 25.1 ± 7.21 % (n = 5; P < 0.01; Fig. 1A), whereas medium doses of DA (10–100 μM) induced either increase or decrease of EPSC amplitudes. In contrast, at high concentration of 200 μM, DA consistently and significantly decreased the NMDA receptor-mediated EPSCs by average of 42.2 ± 5.29 % (n = 5; P < 0.01; Fig. 1A and 1B). This suppressing effect on NMDA currents was long-lasting with little recovery even after 15–20 min washout, suggesting the possibility of NMDA receptor internalization under this condition. This dose-dependent bidirectional effect is in agreement with previous findings that puff NMDA-induced current is dose-dependently regulated by DA (Zheng et al., 1999).
Numerous studies have been conducted involving low-dose DA or D1 agonists (Cepeda and Levine, 1998; Zheng et al., 1999; Seamans et al., 2001); therefore we attempted to identify which DA receptor was mainly activated by high-dose DA to suppress the NMDA receptor-mediated EPSCs. Selective D2 antagonist L741,626 (10 μM) or D1 antagonist SCH 23390 (10 μM) was first applied for 5 minutes and then was co-applied with DA (200 μM). Under this condition, we found that neither L741,626 nor SCH 23390 exhibited significant effects on the NMDA receptor-mediated EPSCs (data not shown), supporting previous reports (Gao et al., 2001; Seamans et al., 2001; Tseng and O’Donnell, 2004). In contrast, L741,626 effectively abolished the suppressing effects of high-dose DA in NMDA receptor-mediated EPSCs in all of the neurons tested (n = 6; P < 0.01; Fig. 2A), in agreement with a recent report in the amygdala (Martina and Bergeron, 2008). Surprisingly, further attenuation of the NMDA receptor-mediated EPSCs was evident when the D1 antagonist SCH 23390 was co-applied with DA (n = 6; P > 0.05; Fig. 2B) (Castro et al., 1999), suggesting that high-dose of DA might activate both D1 and D2 receptors. The effect of D1 activation on NMDA receptors might be opposite to the effect produced by D2 activation. However, under our recording conditions the D2 effect was much stronger and completely dominated the D1 effect when both receptors were co-activated.
Recent studies conducted in the striatal (Dunah and Standaert, 2001; Dunah et al., 2004; Hallett et al., 2006; Tong and Gibb, 2008) and prefrontal neurons (Gao and Wolf, 2008) (Li et al, unpublished observations) indicated that D1 receptor activation led to rapid insertion of NMDA receptors. We therefore attempted to examine whether high-dose of DA-induced depression of NMDA currents also depends on the trafficking, e.g., internalization/endocytosis, of NMDA receptors. It has been reported that young prefrontal neurons, similar to other cortical neurons (Kumar and Huguenard, 2003; Liu et al., 2004; Harris and Pettit, 2007), contain a relatively high proportion of NR2B subunits (Wang et al., 2008). Because NR2B subunits exhibit significantly higher surface mobility and endocytosis than other NMDA receptor subunits (Roche et al., 2001; Lavezzari et al., 2004; Groc et al., 2006), we examined whether high-dose of DA regulates NMDA receptor-mediated EPSCs mainly through NR2B containing NMDA receptors. Selective NR2B subunit antagonist ifenprodil, which preferentially inhibits NR1/NR2B channels with high affinity (IC50 = 0.34 μM) (Williams, 1993; Tovar and Westbrook, 1999), was first applied to block NR2B component in NMDA receptor-mediated EPSCs. As shown in Figure 3A, ifenprodil (3 μM) significantly attenuated the EPSC amplitudes by average of 41.7 ± 3.73 % (n = 6, P < 0.01). The NR2B component in the evoked NMDA-EPSC appeared to be lower than that in monosynaptic connections between layer 5 pyramidal neurons (Wang et al., 2008) but was consistent with a previous study reported in primate PFC layer 2/3 pyramidal neurons (Gonzalez-Burgos et al., 2008) and our own findings (Wang et al, unpublished observation). In fact, there are likely laminar (Wang et al., 2008) or pathway-specific (Kumar and Huguenard, 2003) differences in subunit composition of synaptic NMDA receptors on pyramidal neurons in neocortex. Nevertheless, when ifenprodil and DA (200 μM) were continuously applied together, the EPSC amplitudes were reduced only about 7 % further (49.1 ± 4.42 %), without significant difference (n = 6, P = 0.15). This data suggests that DA mainly interacts with NR2B subunits to achieve its depressive effects in the NMDA receptor-mediated EPSCs, although a lesser effect on NR2A or NR1 subunits cannot be excluded.
Previous studies have suggested a dynamin-dependent endocytosis of ionotropic glutamate receptors (Carroll et al., 1999; Newpher and Ehlers, 2008), particularly NMDA receptors (Nong et al., 2004; Montgomery et al., 2005). We therefore tested whether dynamin-dependent endocytosis of NMDA receptor trafficking underlies DA/D2 receptor-mediated down-regulation of NMDA receptor currents in the PFC neurons. The dynamin inhibitory peptide (QVPSRPNRAP, Biomatik Corporation, Wilmington, DE, USA; 50 μM in pipette) was intracellularly applied to prevent the receptor endocytosis. We found that dynamin inhibitory peptide significantly attenuated the depressing effect of DA on NMDA receptor-mediated EPSCs by about 50 % (n = 6; P < 0.05, Fig. 3B), consistent with a previous study (Montgomery et al., 2005). To investigate endocytosis further, we have tested the effects of another potent membrane permeable inhibitor of endocytosis, dynasore (Tocris Bioscience, Ellisville, Missouri, USA), on DA-induced inhibition of NMDA receptors in the brain slices (Macia et al., 2006; Kirchhausen et al., 2008). Similarly, we found that dynasore (100 μM in bath solution) also blocked about 50% of the inhibition induced by DA in NMDA receptor-mediated currents (n = 6, P < 0.05; Fig. 3B and C). These data indicated that the inhibition of dynamin blocked the NMDA receptor internalization induced by DA, and that the dynamin-dependent endocytosis of NMDA receptors plays an important role in DA-mediated suppression of NMDA responses in the prefrontal neurons (Tong and Gibb, 2008).
Several recent studies have indicated that D2 receptor-mediated GSK-3β signaling pathway is highly associated with hyperdopamine-dependent behaviors (Beaulieu et al., 2005; Beaulieu et al., 2007a). In addition, other evidence suggested that elevated DA may couple to Gq11 signaling pathway by activating both D1 and D2 receptors in the striatal neurons (Lee et al., 2004; Rashid et al., 2007a; Rashid et al., 2007b). Because both D1 and D2 receptors are distributed in same neurons in the PFC, we tested the involvement of these two signaling pathways in the high-dose DA-mediated regulation of NMDA receptors. As shown in Figure 4, co-application of DA (200 μM) with either of the two structurally different GSK-3β inhibitors, TDZD (10 μM) or SB216,763 (10 μM), completely abolished the DA-induced suppression in NMDA receptor mediated-EPSCs (n = 6, P < 0.01 for both). Both inhibitors themselves, however, did not show clear effects on the basal NMDA receptor-mediated transmissions (data not shown), as was reported in previous studies (Peineau et al., 2007; Zhu et al., 2007; Peineau et al., 2008). In contrast, neither Gq11 inhibitor peptide [D-Trp7,9,10] Substance P (Arg-Pro-Lys-Pro-Gln-Gln-Trp-Phe-D-Trp-D-Trp-Met-NH2; Tocris Bioscience, Ellisville, Missouri, USA, 1 μM in pipette), the only inhibitor available for Gq11 blockade (Fig. 4B), nor PLC inhibitor U73122 (10 μM bath application, data not shown) affected the inhibitory effects of DA on the amplitude of the NMDA receptor-mediated EPSCs (n = 6; P > 0.05 for both). This negative effect was consistent with a previous study reported in synaptic NMDA-EPSC in lateral amygdala slices (Martina and Bergeron, 2008) but differs from that in puff-NMDA-induced current (Kotecha et al., 2002).
Since the signaling complex containing protein phosphatase 2 (PP2A) directly mediates the action of D2-class dopamine receptors (Beaulieu et al., 2005) and PP2A is identified as an upstream regulator of GSK-3β (Beaulieu et al., 2009), we tested the effects of PP2A inhibitor okadaic acid (Tocris Bioscience, Ellisville, Missouri, USA) on DA-induced inhibition of NMDA receptor-mediated EPSCs. As shown in Fig. 4C and D, okadaic acid (100 nM) significantly blocked the DA effects (n = 6, P < 0.05). We also examined the role of PI3-kinase, another upstream regulator of GSK-3β (Habas et al., 2006), in NMDA receptor-mediated EPSCs. Bath application of PI3K inhibitor LY294,002 (20 μM, Tocris) only briefly reduced ~50% of the DA-induced suppression in NMDA receptor mediated-EPSCs (19.9 ± 1.82 % compared with that of 40.1 ± 3.95 % in DA alone, n = 6, P < 0.05, data not shown) for about 5 minutes without continuous effects on the DA action (P > 0.05). These results suggest that the PP2A/GSK-3β, but not the Gq11/PLC, signaling pathway is likely involved in the high-dose DA/D2-induced depression of NMDA receptor-mediated EPSCs in the prefrontal neurons. The role of PI3-kinase in DA action, however, is limited.
To investigate whether high concentration of DA also exhibits similar effects in NMDA receptor functions in vivo, GBR12909, a selective DA reuptake inhibitor, was systemically administered to animals to induce acute hyperdopamine state. A single dose of GBR12909 (10 mg/kg, i.p.) was injected with saline as vehicle control and the rats were sacrificed for physiologic recording after 1 hour. The 10 mg/kg of GBR12909 is a widely accepted medium dose (range 2.5 to 40 mg/kg) considered to be very effective in elevating DA levels in the brain (Zahniser et al., 1999; Camarero et al., 2002; Thakker et al., 2004). We found that GBR12909 significantly decreased both the frequency and amplitude of the spontaneous NMDA receptor-mediated EPSCs in the layer 5 pyramidal neurons compared with those in control (n = 6; P < 0.01; Fig. 5A to C). These effects, however, were abolished by pre-treatment with either D2 receptor antagonist L741,626 or GSK-3β inhibitor TDZD. L741,626 is a potent and selective D2 receptor antagonist and is biologically active following systemic administration in vivo. As shown in Figure 5, when L741,626 was injected (10 mg/kg, i.p.) 30 minutes prior to GBR12909 administration, no significant effects were detected, neither for frequency nor for amplitude, on the spontaneous NMDA receptor-mediated EPSCs (n = 6; P > 0.05; Fig. 5). These data indicate that the suppressing effect of GBR12909 likely acts through the activation of D2 receptors (Caruana et al., 2006). Similarly, pretreatment with GSK-3β inhibitor TDZD (1 mg/kg, i.p., 30 minutes before GBR12909 administration) also completely blocked the suppressing effects of GBR12909 on spontaneous NMDA EPSCs. Both frequency and amplitude of the spontaneous EPSCs were recovered to the control levels (n = 6; P > 0.05; Fig. 5). It has been reported that in the frontal cortex, DA might be at least partially, if not mainly, cleared by the norepinephrine transporter (NET) (Moron et al., 2002; Miner et al., 2003; Carboni and Silvagni, 2004; Williams and Steketee, 2004). We have therefore tested the effects of a selective NET inhibitior nisoxetine (10 mg/kg, i.p., Tocris) and a non-selective DA and NE reuptake blocker bupropion (10 mg/kg, i.p., Tocris) on NMDA currents. As shown in Fig. 5D, the selective NET inhibitor nisoxetine exhibited no effect on EPSC frequency but increased the amplitude of NMDA current although not statistically significant (n = 6, P = 0.93 for frequency and P = 0.06 for amplitude). In contrast, bupropion significantly decreased EPSC frequency (n = 6, P < 0.05), similar to GBR12909, but had no significant effects on the EPSC amplitude (n = 6, P = 0.60). The role of NET inhibitor on NMDA receptors in the neocortex remains unclear but seemed to be opposite from that of DA transporter inhibitor. In fact, bupropion acts as not only a norepinephrine and dopamine reuptake inhibitor, but also a nicotinic antagonist (Slemmer et al., 2000). These results further confirmed the effects of GSK-3β on hyperdopamine-induced depression of NMDA receptors, consistent with the finding in PFC slice recording described above. However, the net effect of NET inhibitor and the interactive effect of NE and DA (dual reuptake blocker) on NMDA EPSCs remain to be explored.
To further examine the role of GSK-3β in hyperdopamine-mediated depression of NMDA receptors, GSK-3β protein levels were measured using Western blotting. As shown in Figure 6, GBR12909 injection (10 mg/kg, i.p., single dose, 1 hour) significantly increased the total protein levels of GSK-3β by ~ 40% compared to saline vehicle control (P < 0.01, Fig. 6A and B). Because GSK-3β is activated via its phosphorylation at Tyr216 and is inactivated by phosphorylation at Ser9 through serine/tyrosine (ser/tyr) kinase, we further tested whether hyperdopamine induced by GBR12909 injection also affects the phosphorylation of GSK-3β. The phosphorylation state of GSK-3β was examined using Western blot with phosphor-specific anti-GSK-3β Ser9 and anti-GSK-3β Tyr 216 antibodies. We found that the levels of GSK-3β Tyr216 in GBR12909 treatment were increased by ~ 30% relative to saline vehicle control levels (P < 0.05). In contrast, the levels of phosphor-GSK-3β Ser9 in GBR12909 treatment remained unchanged compared to saline vehicle control levels (P = 0.67) and were significantly lower than total protein level of GSK-3β when the expression levels of phosphor-GSK-3β Ser9 in the GBR12909 treatment were normalized to total GSK-3β levels (P < 0.01). These results suggest an overall increase of GSK-3β activity in the GBR12909-induced hyperdopamine state, consistent with previous reports (Beaulieu et al., 2007a).
Because high-dose of DA decreases the NMDA receptor-mediated EPSCs in the prefrontal cortical slices by inducing internalization of NMDA receptors, largely NR2B subunits, we proposed that the surface expression of NMDA receptor proteins might be affected by hyperdopamine induced by systemic administration of GBR12909 (10 mg/kg, i.p., single dose, 1 hour). We tested this possibility by incubating the prefrontal cortical slices with membrane impermeable crosslinking reagent BS3. Crosslinking assay has been used previously to measure the surface expression of glutamate receptors in brain slices (Grosshans et al., 2001; Conrad et al., 2008). These studies have shown that incubation of brain slices with BS3 does not affect intracellular proteins such as β-actin. Thus, this technique enabled us to identify the relative changes of both surface and intracellular components of NMDA receptors. Here we focused on the NR2B subunits because our in vitro slice physiology experiments (Fig. 3) suggested that NR2B subunits were mainly affected by DA. Our analysis revealed that GBR12909 injection (10 mg/kg, i.p., 1 hour) clearly decreased the protein expressions of both surface and intracellular NR2B subunits. Both the BS3 crosslinked surface NR2B (~500 KD) and the presumptive intracellular NR2B subunits (180 KD) were significantly decreased by GBR12909 administration (Fig. 7A and B). In contrast, β-actin, which is not present on the cell surface, was not affected by BS3 crosslinking reagent in both control and GBR12909-treated rats, indicating the reliability of the BS3 crosslinking procedure (Fig. 7A). Overall, we found that both surface and intracellular expressions of NR2B subunits, as well as the total NR2B protein levels, were significantly decreased (P < 0.05, Fig. 7B). These results indicate that GBR12909 significantly affected the trafficking of NMDA receptors, likely through a process involving inhibition of protein synthesis. This assumption was further confirmed by the evidenced significant decrease in both total protein level (P < 0.05; Fig. 7C) and mRNA expression (P < 0.05; Fig. 7D) of NR2B subunits in GBR12909-treated rat PFC. As shown in Fig. 7C, GBR12909 (10 mg/kg, i.p., 1 hour) significantly decreased the total protein expression of NR2B (P < 0.05) but not NR2A (P > 0.05) detected with Western blot. Importantly, consistent with the electrophysiological changes in prefrontal cortical slices, pretreatment with GSK-3β inhibitor TDZD 30 minutes in advance = (1 mg/kg, i.p.) completely reversed the decrease of protein expressions of both cell surface and intracellular NR2B subunits, the total protein level, as well as the mRNA expression of NR2B (P > 0.05; Fig. 7B to D). These data further suggest that activation of GSK-3β is required for hyperdopaminergic regulation of NMDA receptor trafficking.
The DA-induced endocytosis of NR2B subunit was further tested by incubating the prefrontal cortical slices with membrane impermeable crosslinking reagent BS3 under conditions of DA incubation with and without dynasore. Both surface and intracellular protein levels of NR2B subunits were significantly decreased by DA treatment (200 μM, 10 min, n = 4, P < 0.05). When DA was co-applied with dynasore (100 μM), the effect of DA on surface, but not intracellular, expression of NR2B, was partially but significantly blocked (n = 4, P < 0.05, Fig. 7E). In addition, we wondered whether NR2B phosphorylation plays a role in the D2-mediated suppression of NMDA receptors. Previous study indicated that phosphorylation of NR2B ser-1480 site disrupts the interaction of NMDARs with PSD-95 and SAP102 and decreases surface NR2B expression (Chung et al., 2004), whereas phosphorylation of NR2B at tyr-1472 was involved in synaptic insertion (Hallett et al., 2006; Gao and Wolf, 2008) (Li et al, unpublished observation). We therefore tested whether the trafficking of NR2B induced by hyperdopamine is also through phosphorylation of ser-1480. As shown in Fig. 7F, GBR12909 injection (10 mg/kg, i.p., 1 hour) significantly increased the expression of NR2B ps-1480 (P < 0.05) and this increase was blocked by pre-injection of GSK-3β inhibitor TDZD 30 minutes in advance (1 mg/kg, i.p.), indicating that phosphorylation of NR2B ser-1480 may also affect the surface NR2B expression. It is therefore possible that dynamin and NR2B ps-1480 jointly regulate the endocytosis of NR2B subunits under condition of D2 receptor activation although it is not clear whether dynamin is involved in surface expression of NR2B mediated by phosphorylation of NR2B ser- 1480.
How does activation of GSK-3β affect the NMDA receptors? Previous studies indicated that the β-catenin-binding site lies close to the GSK-3β-binding site and that S33, S37, and T41 sites in β-catenin are strongly phosphorylated by GSK-3β (Daugherty and Gottardi, 2007; Xu and Kimelman, 2007). β-Catenin is known to be localized in postsynaptic density as a component of NMDA receptor multiprotein complex (Husi et al., 2000; Okabe et al., 2003; Al-Hallaq et al., 2007). GSK-3-mediated phosphorylation of β-catenin could allow β-catenin to go to the cell nucleus, interact with transcription factors, and thus regulate gene transcription (Aberle et al., 1997; Xu and Kimelman, 2007). We therefore speculated that hyperdopamine-induced decrease in expression of NR2B receptor protein and mRNA might be caused by regulating β-catenin through activation of GSK-3β. To test this possibility, we first examined the interaction between β-catenin and NR2B with immunoprecipitation. β-catenin was immunoprecipitated with anti-β-catenin and was then probed for immunoprecipitated NR2B with specific anti-NR2B antibody. Under control conditions, we found that NR2B co-precipitated with β-catenin (Fig. 8A), whereas the negative control β-tubulin was not detected in β-catenin immunoprecipitates (data not shown), suggesting a potential protein interaction between β-catenin and NR2B (Husi et al., 2000; Okabe et al., 2003; Al-Hallaq et al., 2007). Consistently, the NR2B co-binding with β-catenin was significantly decreased by GBR12909 administration (10 mg/kg, i.p.) (P < 0.01, Fig. 8A). The total protein level of β-catenin was, however, not changed by GBR12909 treatment, in both immunoprecipitation (P> 0.05, Fig. 8A) and Western blot (P> 0.05, Fig. 8B). The decreased binding between NR2B and β-catenin, however, might also be partially attributable to the decreased expression of NR2B subunits. How does the interaction between NR2B and β-catenin regulate NR2B expression? Although total protein of β-catenin was not significantly afftected by GBR12909 administration, it is possible that β-catenin was phosphorylated and thus affected the NR2B expression by interfering the mRNA expression of this subunit (Schmeisser et al., 2009). Indeed, the expression of phosphor- β-catenin detected with antibody against S33, S37, and T41 sites was significantly increased by GBR12909 treatment (P < 0.05, Fig. 8B). These results thus suggest that phosphorylation of β-catenin mediates the action of GSK-3β in protein synthesis of NMDA receptors.
In recent years, there have been considerable efforts to understand the cellular mechanisms of DA effects on prefrontal neuronal communications. It is generally believed that DA induces a bidirectional regulation in synaptically evoked NMDA receptor-mediated EPSCs in the cortical neurons (Seamans and Yang, 2004; Cepeda and Levine, 2006; Castner and Williams, 2007). A large body of evidence indicates that activation of D1 receptors can antagonistically modulate NMDA responses via intracellular signaling cascades (Snyder et al., 1998; Flores-Hernandez et al., 2002) or direct protein–protein interactions (Lee et al., 2002; Scott et al., 2002; Pei et al., 2004; Scott et al., 2006). The potentiation of NMDA receptor-mediated currents by low doses of DA or D1 agonists is largely through the activation of D1/protein kinase A (PKA)/dopamine and cAMP regulated phosphoprotein 32KD (DARPP32) intracellular pathways (Cepeda and Levine, 1998; Zheng et al., 1999; Greengard, 2001; Seamans et al., 2001; Yang and Chen, 2005). The relationship between D1 stimulation and NMDA-EPSC changes evidenced at individual neurons in vitro is similar to the inverted “U” shape function for D1 receptor stimulation required for in vivo working memory performance (Zahrt et al., 1997; Floresco and Phillips, 2001; Seamans and Yang, 2004; Williams and Castner, 2006; Vijayraghavan et al., 2007). This biphasic/bidirectional modulation of NMDA receptors by D1 stimulation could therefore have important implications given the important role of NMDA receptors in the cellular aspects of working memory (Li et al., 1997; Lisman et al., 1998; Romanides et al., 1999; Durstewitz et al., 2000; Wang et al., 2008).
Despite the extensive study of D1-NMDA receptor interaction, the molecular mechanisms associated with the influences of hyperdopamine on glutamatergic systems, particularly NMDA receptors in the PFC, remain unclear. In this study, we show how elevated DA affects the NMDA receptor functions in the PFC by both in vitro and in vivo approaches. We confirmed that NMDA-EPSCs were potentiated by low doses of DA (≤ 10 μM) but significantly attenuated by higher doses of DA (≥ 100 μM), consistent with previous reports in the rat PFC (Law-Tho et al., 1994; Zheng et al., 1999; Seamans et al., 2001), striatum (Cepeda and Levine, 1998; Jiao et al., 2007), and amygdala (Martina and Bergeron, 2008). Importantly, we found that, under both in vitro and in vivo conditions, the synaptically evoked NMDA receptor-mediated currents in the rat prefrontal neurons were significantly attenuated by high concentration of DA stimulation through the activation of D2 receptors. The high-dose DA-induced inhibition of the NMDA-EPSCs recorded in the prefrontal slices was long-lasting with little recovery. How would D2 activation inhibit the NMDA-EPSCs in the prefrontal neurons? Although presynaptic D2 receptors have been implicated in suppressing glutamate synaptic transmission (Levine et al., 1996; Chen and Gurling, 1999), we, however, think that the D2-mediated inhibition of NMDA receptors was likely through postsynaptic mechanisms, such as receptor internalization/endocytosis. Indeed, high-dose DA failed to affect the NMDA-EPSCs after blockade of NR2B subunits. Intracellularly loaded inhibitory peptide of dynamin, a GTPase responsible for endocytosis, also significantly reduced the DA-induced inhibition of NMDA receptors. In addition, recent studies conducted on cultured prefrontal neurons reported a direct inhibition of puff-NMDA-induced current by activation of D2 or D4 receptors, via either platelet-derived growth factor (PDGF) receptor (Beazely et al., 2006) or PKA-protein phosphatase-1 (PP-1) pathway (Wang et al., 2003). At very high concentration of DA (0.1–3 mM), a rapidly reversible and voltage-dependent blockade of inward NMDA-induced current had also been shown in cultured hippocampal neurons (Castro et al., 1999; Kotecha et al., 2002). However, whether the signaling pathways reported in cultured neurons are analogous to those in vivo remain to be determined. In fact, there is no PDGF receptor expression in the postnatal cortical neurons (Ellison et al., 1996).
Recent studies indicated that elevation of DA levels induced by administration of DA transporter inhibitors such as cocaine and amphetamine resulted in cognitive disorders via D2-GSK-3β pathway (Beaulieu et al., 2005; Beaulieu et al., 2007a; Lute et al., 2008; Beaulieu et al., 2009). The proposed role of GSK-3β in DA/D2 receptor- associated psychiatric disorders is intriguing, yet whether and how it will affect NMDA receptor-mediated synaptic transmission in vivo remains untested. Here we provided the first evidence that hyperdopamine induced by DA reuptake inhibitor GBR12909 significantly decreases the expressions of both surface and intracellular NR2B proteins, as well as mRNA levels of NR2B, in the prefrontal cortical neurons, likely through the inhibition of protein synthesis. Furthermore, the high-dose DA/D2-mediated suppression of NMDA receptors required the activation of GSK-3β through increasing the phosphorylation of GSK-3β Tyr216 and decreasing the phosphorylation of GSK-3β Ser9, but neither dependent on Gq11 nor PLC signaling. Our data support several recent studies, which suggested that activation of GSK-3β is associated with NMDA receptor-dependent synaptic plasticity (Peineau et al., 2007; Zhu et al., 2007; Peineau et al., 2008). Numerous studies have emphasized the importance of GSK-3β in NMDA receptor-mediated neuroprotection and neurodegeneration (Takadera et al., 2004; Szatmari et al., 2005; Habas et al., 2006; Hetman and Kharebava, 2006; Lei et al., 2008), as well as synaptic plasticity (Hooper et al., 2007; Peineau et al., 2007; Zhu et al., 2007; Peineau et al., 2008). Previous studies indicated that phosphorylation of GSK-3β Tyr216 is required for basal activity and increased phosphorylation of this residue result in increased activity of GSK-3β (Hughes et al., 1993), whereas phosphorylation of GSK-3β Ser9 leads to inactivation of GSK-3β, overriding the activation induced by phosphorylation of GSK-3β Tyr216 (Bhat et al., 2000). Based on our results, we therefore conclude that an increase in GSK-3β and GSK-3β Tyr216 levels together with the relatively lower expression of GSK-3β Ser9 are consistent with an overall increase of GSK-3β activity in the GBR12909-induced hyperdopaminergic state. The increased GSK-3β-mediated inhibition of NMDA receptors makes an ideal case for its role in the induction of long-term depression (Peineau et al., 2007; Peineau et al., 2008) and inhibition of long-term potentiation (Zhu et al., 2007) because both synaptic processes are NMDA receptor-dependent. Our result is also consistent with previous studies conducted in both in vivo and in vitro conditions, indicating that GSK-3β inhibitors had no clear effects on the basal NMDA receptor-mediated synaptic transmission (Peineau et al., 2007; Zhu et al., 2007; Peineau et al., 2008). In another study using cultured neurons, however, GSK-3β inhibitor was found to slightly inhibit NMDA-induced current (Chen et al., 2007). It is likely that the discrepancy derived from the different neuronal preparations, and the dissociated neurons might be more sensitive to GSK-3β inhibitors.
Previous studies indicated that activation of GSK-3β can induce phosphorylation and thus degradation of β-catenin (Liu et al., 2002; Daugherty and Gottardi, 2007). We, however, found that the total protein level of β-catenin was unaltered. This seemingly contradicts previous studies. Instead, we did observe increased phosphorylation of β-catenin S33/37/T41. It is possible that increase of GSK-3b activity induces phosphorylation of β-catenin that in turn results in the redistribution of β-catenin (Murase et al., 2002) instead of degradation (Liu et al., 2002; Daugherty and Gottardi, 2007). Another possibility is that degradation of β-catenin induced by phosphorylation occurred beyond of test time applied in our experiments, which was one hour after GBA12909 administration. This proposition, however, needs further exploration. Nevertheless, our data suggest that GSK-3β plays an essential role in regulating NMDA receptor functions in the prefrontal neurons. Based on our findings, we speculate that GSK-3β is likely activated by its upstream regulator PP2A (Beaulieu et al., 2009) and it in turn phosphorylates β-catenin which interacts with NR2B, controls the NR2B gene transcription, and thus affects the protein synthesis of NR2B subunits (Daugherty and Gottardi, 2007; Xu and Kimelman, 2007). Consequently, the NR2B surface expression and NMDA receptor functions are regulated by D2 receptor activation (Fig. 9).
Although concurrent changes of DA and NMDA receptors play an essential role in the pathogenesis of psychiatric disorders such as schizophrenia, little is known about the correlation between hyperfunction of the dopaminergic system and NMDA receptor functions in vivo. In this study, we present evidence that an animal model with persistently elevated dopamine levels is ideal for studying the action of hyperdopamine on NMDA receptor functions. We administered DA reuptake inhibitor GBR12909 to induce acute hyperdopamine state in rats (Carlsson et al., 1999). Numerous studies have shown that this potent and highly selective DA reuptake inhibitor can cause significant elevation of DA levels in the extracellular space (Karoum et al., 1994; Zahniser et al., 1999; Easton et al., 2007). In addition, GBR12909 selectively increases DA levels without affecting norepinephrine and serotonin because the affinity for the latter two transmitters is 100-fold lower (Andersen, 1989). The neurochemical and behavioral consequences of this blockade are similar to those obtained using DA transporter gene knockdown (Zhuang et al., 2001; Thakker et al., 2004; Cagniard et al., 2006) and thus it serves as a good model to study the hyperdopaminergic state. The increased DA neurotransmission arising from administration of GBR12909 results in the activation of D2 receptors, which in turn induces inhibition of NMDA receptors, particularly via NR2B subunit endocytosis and phosphorylation, through the constant activation of GSK-3β. These data support the involvement of GSK-3 as an important mediator of DA action in vivo and suggest that modulation of the PP2A-GSK-3β pathway might be relevant to DA-related disorders, such as schizophrenia and drug addiction (Beaulieu et al., 2007a; Koros and Dorner-Ciossek, 2007; Beaulieu et al., 2009). In summary, our study suggests that GSK-3β signaling pathway is required for the hyperdopamine/D2 receptor- mediated inhibition of NMDA receptors in the prefrontal neurons and this action may underlie D2 receptor-mediated psychostimulant effects and hyperdopamine-dependent behaviors in the brain.
We are grateful to Drs. Feng Yang and Victoria A Zhukareva for their comments on the manuscript. Anti-NR2B-pS1480 antibody was gifted from Dr. Richard L. Huganir. This study was supported by research fund of Drexel University College of Medicine, a grant from a NARSAD young investigator award, NIH R21MH232307 and R01MH085666 to W-J Gao; and research fund of Drexel University College of Medicine, NARSAD young investigator award (2005) to Y-Q Huang. The NIH had no further role in the study design; in the collection, analysis, and interpretation of the data; in the writing of the report; and in the decision to submit the paper for publication.
Conflict of Interest: The authors claim no conflict of interest.
Author ContributionsW.J.G. conceived the study, supervised the project, and wrote the manuscript. Y.C.L carried out the experiments and data analysis of electrophysiology, Western blots, immunoprecipitation, and wrote the manuscript. D.X. and J.R. carried out the experiments of PCR and Western blots. Y.Q.H. supervised the biochemical experiments and contributed to the manuscript preparation.