A single in vivo cocaine administration impairs 5-HT1B receptor-induced long-term depression in the nucleus accumbens

Authors


Address correspondence and reprint requests to Kuei-Sen Hsu, Ph.D., Department of Pharmacology, College of Medicine, National Cheng Kung University, No. 1, University Rd., Tainan 701, Taiwan. E-mail: richard@mail.ncku.edu.tw

Abstract

The nucleus accumbens (NAc) is a crucial forebrain nucleus implicated in reward-based decision-making. While NAc neurons are richly innervated by serotonergic fibers, information on the functional role of serotonin 5-hydroxytryptamine (5-HT) in the NAc is still sparse. Here, we demonstrate that brief application of 5-HT or 5-HT1B receptor agonist CP 93129 induced a long-term depression (LTD) of glutamatergic transmission in NAc neurons. This LTD was presynaptically mediated and inducible by endogenous 5-HT. Remarkably, a single cocaine exposure impaired the induction of LTD by 5-HT or CP 93129. The inhibition was blocked when a selective dopamine D1 receptor antagonist SCH23390 was coadministered with cocaine. Cocaine treatment resulted in increased phosphorylation of presynaptic proteins, rabphilin 3A and synapsin 1, and significantly attenuated CP 93129-induced decrease in rabphilin 3A and synapsin 1 phosphorylation. Application of cAMP-dependent protein kinase inhibitor KT5720 caused a prominent synaptic depression in NAc neurons of mice with a history of cocaine exposure. Our results reveal a novel 5-HT1B receptor-mediated LTD in the NAc and suggest that cocaine exposure may result in elevated phosphorylation of presynaptic proteins involved in regulating glutamate release, which counteracts the presynaptic depressant effects of 5-HT1B receptors and thereby impairs the induction of LTD by 5-HT.

Abbreviations used
5-HT

5-hydroxytryptamine

ACSF

artificial cerebrospinal fluid

AMPA

2-Amino-3-(3-hydroxy-5-methyl-isoxazol-4-yl)propanoic acid

CNQX

6-cyano-7-nitroquinoxaline-2,3-dione

DA

dopamine

D-APV

D-2-amino-5-phosphonovalerate

DMSO

dimethyl sulfoxide

ECL

enhanced chemiluminescence

EPSCs

excitatory postsynaptic currents

HRP

horseradish peroxidase

LTD

long-term depression

MSNs

medium spiny neurons

NAc

nucleus accumbens

NAN-190

1-(2-methoxyphenyl)-4-(4-phthalimidobutyl)piperazine

PBS

phosphate-buffered saline

PKA

protein kinase

PPR

paired-pulse ratio

SDS-PAGE

sodium dodecyl sulfate–polyacrylamide gel electrophoresis

VTA

ventral tegmental area

The nucleus accumbens (NAc), a central component of the mesolimbic dopamine system, has been implicated in an array of functions ranging from motivation and reward to feeding and drug addiction (Hyman et al. 2006; Thomas et al. 2008). The GABAergic medium spiny neurons (MSNs) constitute the major cell type of the NAc (> 95% of total) (Smith and Bolam 1990; Sesack and Grace 2010). Besides receiving dopaminergic afferents from the ventral tegmental area (VTA) and glutamatergic inputs from the limbic system such as the prefrontal cortex, hippocampus and amygdala (Kalivas 2004), NAc MSNs are also richly innervated by serotonergic 5-hydroxytryptamine (5-HT) fibers originating from the dorsal raphe nucleus of the midbrain (Steinbusch 1981; Phelix and Broderick 1995; Van Bockstaele et al. 1996). Recent studies emphasize the importance of dopaminergic and glutamatergic transmission in the NAc in mediating reward-based learning and motivation; however, little is known about the functions of the NAc 5-HT system. Interestingly, a previous study has shown that 5-HT can reversibly inhibit glutamatergic transmission in adult rat NAc neurons through activation of presynaptic 5-HT1B receptors (Muramatsu et al. 1998). However, a persistent presynaptic inhibition of glutamatergic transmission in the NAc core by the 5-HT1B receptor agonist has been recently reported in young rats (Mathur et al. 2011), suggesting that the extent and duration of 5-HT1B receptor activation, and coupling to downstream signaling pathways, may be important in governing the transition from short- to long-term inhibition of glutamate release. However, the functional relevance of the 5-HT1B receptor-mediated synaptic depression in these studies is not clear.

In addition to the well-recognized dopaminergic mechanisms in the development of cocaine addiction, an increasing number of evidence indicates that cocaine-induced alterations in the NAc 5-HT system may also contribute to behavioral sequelae of addiction. For example, there is evidence that elevated expression or activation of 5-HT1B receptors in NAc neurons that project to the VTA can promote the development of cocaine-induced behavioral sensitization (Neumaier et al. 2002; Przegaliński et al. 2002). In addition, a recent study has revealed that increased activation of 5-HT1B receptors in NAc shell neurons can enhance both the rewarding and aversive properties of cocaine (Barot et al. 2007). Given cocaine-induced long-lasting adaptations of glutamatergic transmission in the NAc are thought to contribute to drug addiction and related behaviors (Schmidt and Pierce 2010; Hyman et al. 2006), it seems reasonable to speculate that dysregulation of serotonergic modulation of glutamatergic transmission in the NAc would have a significant role in the development of cocaine addiction. To test this hypothesis, we examined the effects of 5-HT on glutamatergic transmission onto MSNs of the NAc in mice following a single exposure to cocaine. Here, we demonstrate a 5-HT-induced long-term depression (5-HT-LTD) of glutamatergic transmission that is triggered by pharmacological activation or synaptic stimulation of presynaptic 5-HT1B receptors in NAc MSNs. More importantly, we also provide the first evidence that a single in vivo exposure to cocaine impairs the induction of 5-HT-LTD in the NAc.

Materials and methods

Animals

All experiments were performed with male C57BL/6 mice (24–28 days old) bred in the Laboratory Animal Center of National Cheng Kung University. Mice were housed in groups of four in a humidity- and temperature-controlled (22°C) vivarium on a 12 h light/dark cycle (light on at 6:00 A.M.) with food and water available ad libitum. After a 3-day acclimation period, animals were assigned randomly to two groups that received a single intraperitoneal injection of either cocaine (15 mg/kg) or saline (10 ml/kg) and were killed 24 h after the treatment. All comparisons between saline- and cocaine-treated groups were performed by experimenters blind to group assignment. All procedures were conducted in adherence to the National Institutes of Health Guide for the Care and Use of Laboratory Animals and were approved by the Institutional Animal Care and Use Committee of National Cheng Kung University. All efforts were made to minimize the number of animals used and their suffering.

Locomotor activity measurement

Open field locomotor activity was measured as described previously (Huang et al. 2011). Following intraperitoneal injection of cocaine (15 mg/kg) or saline (10 ml/kg), mice were immediately placed in the activity chamber (45 × 20 × 20 cm3) and horizontal locomotor activity was recorded with a video tracking system (Ethovision; Noldus, The Netherlands) for 20 min under dim light, sound-attenuated conditions. Distance traveled was analyzed for estimates of locomotor activity. For locomotor habituation to the activity chamber, mice were placed in the chamber 20 min/day for three consecutive days. Following habituation, all mice received 2 days of saline injections and then were divided into two groups that received injections of cocaine or saline (during 10:00–12:00 h). To assess the role of 5-HT1B receptors in cocaine-induced hyperlocomotion, the 5-HT1B receptor antagonist (2R)-2-[[[3-(4-morpholinylmethyl)-2H-1-benzopyran-8-yl]oxy]methyl]morpholine dimethanesulfonate (NAS-181; 3 mg/kg) was administered intraperitoneally 15 min before the injection of saline or cocaine.

Acute slice preparation and electrophysiology

Acute coronal slices and whole-cell patch-clamp recordings were conducted as described previously (Huang et al. 2011; Huang and Hsu 2012). In brief, mice were anesthetized with isoflurane (3%–4%, inhaled; Halocarbon laboratories, River Edge, NJ, USA) and decapitated with guillotine, and their brains were rapidly removed and placed in ice-cold sucrose artificial cerebrospinal fluid (ACSF) cutting solution [containing (in mM): sucrose 234, KCl 2.5, CaCl2 0.5, MgCl2 7, NaHCO3 25, NaH2PO4 1.25, and glucose 11 at pH 7.3–7.4 and equilibrated with 95% O2–5% CO2]. Slices (250 μm) were prepared using a vibrating microtome (VT1200S; Leica, Nussloch, Germany) and transferred to a holding chamber of normal ACSF [containing (in mM): NaCl 117, KCl 4.7, CaCl2 2.5, MgCl2 1.2, NaHCO3 25, NaH2PO4 1.2, and glucose 11 at pH 7.3–7.4 and equilibrated with 95% O2–5% CO2] and maintained at 25°C for at least 1 h before use.

Whole-cell patch-clamp recordings were performed from visualized MSNs located in the NAc core. The MSNs were voltage-clamped at −70 mV and recordings were made using a patch-clamp amplifier (Axopatch 200B, Axon Instruments, Union City, CA, USA). An Intel Pentium-based computer with pCLAMP software (Version 8.0; Axon Instruments) was used for on-line acquisition and off-line analysis of the data. For measurement of synaptically evoked excitatory postsynaptic currents (EPSCs), a bipolar stainless steel stimulating electrode was placed 150–200 μm rostral to the recording electrode to stimulate excitatory afferents at 0.05 Hz and the superfusate routinely contained gabazine (SR95531, 10 μM) to block GABAA receptor-mediated inhibitory synaptic responses. 2-Amino-3-(3-hydroxy-5-methyl-isoxazol-4-yl)propanoic acid (AMPA) receptor-mediated excitatory postsynaptic current (EPSCAMPA) was recorded in the presence of D-2-amino-5-phosphonovalerate (D-APV, 50 μM) and gabazine (10 μM). N-methyl-D-aspartate (NMDA) receptor-mediated EPSC (EPSCNMDA) was recorded in Mg2+-free ACSF containing 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX, 20 μM) and gabazine (10 μM). The paired-pulse ratio (PPR) was calculated as the mean peak amplitude in response to the second stimulus divided by the mean peak amplitude of the response in response to the first stimulus. In some experiments, LTD was induced by a burst stimulation protocol consisting of six pulses (50 ms interpulse interval) every 20 s for 10 min in the presence of citalopram (4 μM) as previously described (Mathur et al. 2011). The composition of intracellular solution was (mM): K-gluconate, 115; KCl, 20; HEPES, 10; MgCl2, 2; EGTA, 0.5; Na2ATP, 3; Na3GTP, 0.3; QX-314, 5 and sucrose to bring osmolarity to 290–300 mOsm/L and pH to 7.3. Series resistance and input resistance were monitored on-line throughout the whole-cell recording with a 5-mV depolarizing step given after every afferent stimulus and data were discarded if access resistance changed by more than 20%.

Miniature EPSCs (mEPSCS) were recorded from MSNs held in voltage clamp at a potential of −70 mV in the presence of gabazine (10 μM), tetrodotoxin (1 μM), and CdCl2 (100 μM) and analyzed off-line using a commercially available software (Mini Analysis 4.3; Synaptosoft, Leonia, NJ, USA). mEPSCs were abolished by CNQX (20 μM) plus D-APV (50 μM), indicating that these are glutamatergic events. The software detects events based on amplitudes exceeding a threshold set just above the baseline noise of the recording (-3 pA).

Western blot

Western blot analysis was performed as described previously (Huang et al. 2011). The microdissected NAc core tissue samples were lysed in ice-cold Tris-HCl buffer solution (TBS; pH 7.4) containing a cocktail of protein phosphatase and proteinase inhibitors to avoid dephosphorylation and degradation of proteins, and ground with a pellet pestle (Kontes glassware, Vineland, NJ, USA). Samples were sonicated and spun down at 15 000× g for 10 min at 4°C. The supernatant was then assayed for total protein concentration using Bio-Rad Bradford Protein Assay Kit (Hercules, CA, USA). Each sample from tissue homogenate was separated in 10% sodium dodecyl sulfate–polyacrylamide gel electrophoresis gel. Following the transfer on nitrocellulose membranes, blots were blocked in buffer solution containing 5% milk and 0.1% Tween-20 in phosphate-buffered saline (124 mM NaCl, 4 mM KCl, 10 mM Na2HPO4, and 10 mM KH2PO4, pH 7.2) for 1 h and then blotted for 2 h at 25°C with the rabbit polyclonal antibodies that recognize 5-HT1B receptors (1 : 1000; Santa Cruz Biotechnology, Santa Cruz, CA, USA), phospho-rabphilin 3A (pS234, 1 : 2000; Abcam, Cambridge, MA, USA), phospho-synapsin 1 (pSer9, 1 : 2000; Cell Signaling Technologies, Beverly, MA, USA) or β-actin (1 : 2000, Sigma-Aldrich, St. Louis, MO, USA). It was then probed with horseradish peroxidase-conjugated secondary antibody for 1 h and developed using the enhanced chemiluminescence immunoblotting detection system (Amersham Biosciences, Buckinghamshire, UK). The immunoblots using phosphorylation site-specific antibodies were subsequently stripped and reprobed with the following antibodies: anti-rabphilin 3A antibody (1 : 1000) that was purchased from abcam and anti-synapsin 1 antibody (1 : 1000) that was obtained from Abcam (Cambridge, MA, USA). Immunoblots were analyzed by densitometry using Bio-profil BioLight PC software (Vulber Lourmat, France). Only film exposures that were not saturated were used for quantification analysis.

Drug treatment

Cocaine HCl (15 mg/kg), (R)-(+)-chloro-8-hydroxy-3-methyl-1-phenyl-2,3,4,5-tetrahydro-H-3-benzazepine (SCH23390, 0.5 mg/kg), raclopride (0.5 mg/kg), and citalopram (10 mg/kg) were dissolved in 0.9% NaCl and administered intraperitoneally. Drug doses were selected on the basis of published studies (Dong et al. 2004; Mateo et al. 2004; Huang et al. 2011). All drugs used in in vitro experiments were applied by manually switching the superfusate. Drugs were diluted from stock solutions just before application. (5R,6S,8S)-Hexyl- 6-hydroxy-5-methyl-13-oxo-6,7,8,13,14,15-hexahydro-5H-16-oxa-4b,8a,14-triaza-5,8-methanodibenzo[b,h]cycloocta[jkl]cyclopenta[e]-as-indacene-6-carboxylate (KT5720) and 1-(2-methoxyphenyl)-4-(4-phthalimidobutyl)piperazine (NAN-190) were dissolved in dimethyl sulfoxide stock solutions and stored at −20°C until the day of experiment. Other drugs used in this study were dissolved in distilled water. 5-HT, 1,4-dihydro-3-(1,2,3,6-tetrahydro-4-pyridinyl)-5H-pyrrol[3,2b]pyridin-5-one dihydrochloride (CP 93129), NAS-181, NAN-190, citalopram, KT5720, SCH23390, raclopride, CNQX, D-APV, and gabazine were purchased from Tocris Cookson (Bristol, UK); cocaine HCl was obtained from Sigma (St Louis, MO, USA).

Statistical analysis

The data for each experiment were normalized relative to baseline, and are presented as means ± SEM. Number of animals used is indicated by n. The significance of the difference between the groups was calculated by one-way anova with Bonferroni's post hoc analyses or Student's t-test where appropriate. Probability values (p) of less than 0.05 were considered to represent significant differences.

Results

5-HT induces LTD via activation of presynaptic 5-HT1B receptors

In the first set of experiments, we examined the effects of 5-HT on glutamatergic transmission onto MSNs of the NAc core. Consistent with previous report (Mathur et al. 2011), bath application of 5-HT (5 μM) for 10 min caused a significant reduction of EPSCs and the synaptic response remained at a depressed level 30 min after 5-HT washout (measured 25–30 min after 5-HT washout: 63.2 ± 6.4% of baseline, n = 10; < 0.05; representative experiment in Fig. 1a and averaged data in Fig. 1b and d). This 5-HT-LTD was completely blocked by a selective 5-HT1B receptor antagonist NAS-181 (20 μM; 93.5 ± 4.7% of baseline, n = 6; < 0.05 compared with 5-HT alone; Fig. 1b and d), but not by the 5-HT1A receptor antagonist NAN-190 (10 μM; Fig. 1d). In parallel, application of 5-HT1B receptor agonist CP 93129 (2 μM) for 10 min mimicked the effect of 5-HT to induce LTD (CP 93129-LTD; measured 25–30 min after CP 93129 washout: 66.5 ± 6.8% of baseline, n = 8; < 0.05; Fig. 1c), which was also blocked by NAS-181 (20 μM; 96.8 ± 3.5% of baseline, n = 5; < 0.05 compared with CP 93129 alone; Fig. 1c and d), but not by the NAN-190 (Fig. 1d). These results suggest that 5-HT1B receptors are responsible for the induction of LTD by 5-HT in the NAc.

Figure 1.

Activation of 5-hydroxytryptamine (5-HT)1B receptors induces long-term depression (LTD) in medium spiny neurons of the nucleus accumbens (NAc). (a) Example of an experiment showing that bath application of 5-HT (5 μM) for 10 min elicited LTD of excitatory postsynaptic currents (EPSCs). (b) Summary of experiments showing that the selective 5-HT1B receptor antagonist NAS-181(20 μM; ●) treatment prevented the induction of 5-HT-LTD. (c) Summary of experiments showing that bath application of 5-HT1B receptor agonist CP 93129 (2 μM; △) for 10 min produced LTD of EPSCs, which was prevented by pretreament of the slices with NAS-181 (20 μM; ▲). (d) Bar chart comparing the effects of NAN-190 (10 μM) and NAS-181 (20 μM) treatment on the induction of LTD by 5-HT and CP 93129. The magnitude of LTD was measured 25–30 min after 5-HT or CP 93129 washout. Representative traces of EPSCs were taken at the time indicated by number. The total number of neurons examined is indicated by n in parenthesis. Horizontal bars denote the period of delivery of drugs. The dashed lines show the level of baseline. *< 0.05 compared with vehicle (0.1% dimethyl sulfoxide) slices. Error bars indicate SEM.

To determine the site of expression of 5-HT-LTD, we examined the effects of CP 93129 on AMPA and NMDA receptor-mediated component of synaptic transmission, respectively. If 5-HT-LTD were expressed presynaptically, changes in the magnitude of both EPSCAMPA and EPSCNMDA by CP 93129 would be expected. In accordance with this view, we found that application of CP 93129 (2 μM) for 10 min induced a nearly identical LTD of AMPA and NMDA receptor-mediated EPSCs. On average, the magnitude of CP 93129-LTD of EPSCAMPA was 27.5 ± 5.3% (n = 5; measured 25–30 min after CP 93129 washout; Fig. 2a). Comparable results were obtained for EPSCNMDA (23.5 ± 5.9%, n = 5; = 0.63; Fig. 2b).

Figure 2.

5-HT-LTD is presynaptically mediated. (a) Summary of experiments showing the effect of CP 93129 (2 μM) on 2-Amino-3-(3-hydroxy-5-methyl-isoxazol-4-yl)propanoic acid receptor-mediated excitatory postsynaptic currents (EPSCs) (EPSCAMPA). EPSCAMPA was recorded in the presence of D-APV (50 μM) and gabazine (10 μM) at a holding potential of −70 mV. (b) Summary of experiments showing the effect of CP 93129 (2 μM) on NMDA receptor-mediated EPSCs (EPSCNMDA). EPSCNMDA was recorded in Mg2+-free ACSF containing CNQX (20 μM) and gabazine (10 μM) at a holding potential of −70 mV. (c) Summary of experiments showing that the paired-pulse ratio was significantly increased after application and then washout of CP 93129 (2 μM). Paired-pulse ratio was calculated from responses to paired-pulse stimulation with an interstimulus interval of 40 ms. (d) Left sample traces (three traces superimposed) of mEPSCs before, during, and after washout of 10-min CP 93129 (2 μM). Right traces are the averaged mEPSCs of 20 events each before, during, and after washout of 10-min CP 93129 with increasing time resolution, demonstrating the lack of effect on the amplitude and kinetics of mEPSCs. (e) and (f) Summary of experiments showing the effects of CP 93129 (2 μM) on the average frequency (e) and amplitude (f) of mEPSCs. Holding potential, −70 mV. The magnitude of changes was measured before (baseline), 10 min after CP 93129 application or 25–30 min after washout. Representative traces of EPSCs in (a) and (b) were taken at the time indicated by number. The total number of neurons examined is indicated by n in parenthesis. *< 0.05 compared with baseline.

To further test the possibility that 5-HT-LTD expresses presynaptically, we examined the effect of CP 93129 (2 μM) on paired-pulse facilitation, a well-defined presynaptic process in which the residual Ca2+ influx following the first of two presynaptic action potentials leads to enhancement of the transmitter release in response to the second action potential (Zucker 1989). We calculated the PPR before and after CP 93129-LTD induction. If the expression of CP 93129-LTD involves a presynaptic mechanism of action, it would be associated with an increase in the PPR. Two consecutive stimuli with a 40-ms interstimulus interval elicited a pair of EPSCs with the second EPSC significantly larger than the first EPSC (Fig. 2c). Under baseline condition, the ratio of the amplitude of second EPSC divided by the first was 1.41 ± 0.06 (n = 9). The expression of CP 93129-LTD was accompanied by a significant increase in PPR to 1.74 ± 0.07 measured 25–30 min after CP 93129 washout (n = 9; < 0.01; paired Student's t-test; Fig. 2c).

We also measured mEPSCs before and after CP 93129-LTD induction. mEPSCs in MSNs were measured under voltage clamp at −70 mV and were pharmacologically isolated from spontaneous inhibitory currents by the inclusion of 10 μM gabazine in the ACSF perfusing the slices. The mEPSCs were totally blocked by bath coapplication of CNQX (20 μM) and D-APV (50 μM), confirming them to be true glutamate receptor-mediated events (data not shown). In eight neurons examined, the expression of CP 93129-LTD was accompanied by a significant decrease in the mean frequency of mEPSCs from 3.16 ± 0.29 Hz to 2.02 ± 0.17 Hz (< 0.01; paired Student's t-test), whereas there was no significant effect of CP 93129 treatment on the mEPSC amplitude (baseline: 11.12 ± 0.63 pA; 25–30 min after CP 93129 washout: 10.68 ± 0.56 pA; = 0.61; paired Student's t-test; representative experiment in Fig. 2d and averaged data in Fig. 2e and f). Together these data provide converging evidence that the locus of NAc 5-HT-LTD expression is presynaptic.

A single in vivo exposure to cocaine impairs 5-HT-LTD induction

Our next approach was to investigate whether the inducibility of 5-HT-LTD in the NAc was modified in the cocaine-treated mice. To evaluate the effect of acute cocaine exposure on 5-HT-LTD induction, we compared the magnitude of LTD induced by 5-HT (5 μM, 10 min) in NAc slices from mice, 24 h following a single saline (10 ml/kg) or cocaine (15 mg/kg) injection. We found that the induction of 5-HT-LTD in the NAc was impaired in slices from cocaine-treated mice (95.3 ± 4.2% of baseline, n = 8; < 0.05; representative experiment in Fig. 3a and averaged data in Fig. 3b and e) compared with those from saline-treated mice (64.5 ± 6.7% of baseline, n = 8). Likewise, mice received an injection of cocaine exhibited impaired CP 93129-LTD (94.2 ± 6.5% of baseline, n = 8; < 0.05; representative experiment in Fig. 3c and averaged data in Fig. 3d and e) compared with those from saline-treated mice (68.7 ± 6.3% of baseline, n = 8). When analyzed as a function of time after cocaine treatment, we found that both 5-HT-LTD and CP 93129-LTD were significantly reduced in cocaine-treated mice for up 12 h–3 days, but not 7–14 days following cocaine exposure, with maximal suppression seen at 24 h (Fig. 3e), suggesting that cocaine-induced modification of 5-HT1B receptor function in the NAc is persistent but not permanent.

Figure 3.

A single in vivo exposure to cocaine impairs the induction of 5-hydroxytryptamine (5-HT)- long-term depression (LTD) in medium spiny neurons of the nucleus accumbens (NAc) core. (a) Example of an experiment showing that bath application of 5-HT (5 μM) for 10 min only elicited a transient depression of excitatory postsynaptic currents (EPSCs) in slice from mouse that had received a single cocaine injection (15 mg/kg, intraperitoneal) 24 h earlier. (b) Summary of experiments showing that a single cocaine exposure impaired the induction of 5-HT-LTD compared with saline-treated controls. (c) Example of an experiment showing that bath application of CP 93129 (2 μM) for 10 min only elicited a transient depression of EPSCs in slice from mouse that had received a single cocaine injection (15 mg/kg, intraperitoneal) 24 h earlier. (d) Summary of experiments showing that a single cocaine exposure (△) impairs the induction of CP 93129-LTD compared with saline-treated controls (▲). (e) Loss of 5-HT- and CP 93129-LTD was analyzed in the NAc at distinct time points after a single cocaine injection. The magnitude of LTD was measured 25–30 min after 5-HT or CP 93129 washout. The total number of animals examined is indicated by n in parenthesis. *< 0.05 compared with saline-treated controls.

The results thus far were obtained with exogenous application of 5-HT or CP 93129, and it is not clear whether the endogenously released 5-HT in the NAc can effectively activate 5-HT1B receptors to elicit LTD. Because bath application of a 5-HT uptake inhibitor citalopram along with a burst stimulation protocol (six pulses at a 50-ms interpulse interval every 20 s for 10 min) at corticostriatal synapses can apparently elicit LTD via activation of 5-HT1B receptors (Mathur et al. 2011), we used this protocol to test our hypothesis. As shown in Fig. 4a, delivery of this burst stimulation protocol in the presence of citalopram (4 μM) consistently produced a stable form of LTD in the NAc (76.5 ± 5.2% of baseline, n = 5; < 0.05), which was completely blocked by NAS-181 (20 μM; 95.3 ± 4.2% of baseline, n = 4), indicating that 5-HT1B receptors mediate this effect. Consistent with 5-HT-LTD observed in cocaine-treated mice, the induction of LTD by burst stimulation plus citalopram was impaired in slices from cocaine-treated mice (25–30 min after citalopram washout: 94.2 ± 3.9% of baseline, n = 5; < 0.05) compared with those from saline-treated mice (78.6 ± 5.9% of baseline, n = 5; Fig. 4b).

Figure 4.

A single in vivo exposure to cocaine impairs the induction of endogenous 5-hydroxytryptamine (5-HT)-long-term depression (LTD) in medium spiny neurons of the nucleus accumbens (NAc) core. (a) Summary of experiments showing that citalopram (4 μM) coupled with the burst stimulation protocol (6 pulses at a 50 ms interpulse interval every 20 s for 10 min; ○) consistently elicited LTD of excitatory postsynaptic currents (EPSCs), which was prevented by pretreament of the slices with NAS-181 (20 μM; ●). (b) Summary of experiments showing that a single cocaine exposure (○) abrogated the induction of LTD by burst stimulation protocol plus citalopram compared with saline-treated controls (●).

A single in vivo exposure to cocaine impairs 5-HT-LTD via the activation of dopamine D1-like receptors

We next investigated how a single in vivo exposure to cocaine alters 5-HT-LTD. Considering that cocaine blocks the dopamine (DA) reuptake transporter and thereby increases local DA concentrations in brain areas receiving dopaminergic inputs (Hyman 1996), it is therefore possible that activation of DA receptors in the NAc is required for the aforementioned cocaine-induced impairment of 5-HT-LTD. To examine the role of DA receptors, we administrated specific DA D1- or D2-like receptor antagonists 15 min before cocaine injection. As shown in Fig. 5a and b, neither the DA D1 receptor antagonist SCH23390 (0.5 mg/kg) nor the DA D2 receptor antagonist raclopride (0.5 mg/kg) had a significant effect on CP 93129-LTD when administered with saline (SCH23390: 67.5 ± 5.3% of baseline, n = 4; raclopride: 69.7 ± 5.8% of baseline, n = 4; > 0.05 compared with saline alone). When coadministered with cocaine, SCH23390, but not raclopride, the induction of CP 93129-LTD was intact in slices from cocaine-treated mice (SCH23390: 75.2 ± 6.4% of baseline, n = 6; raclopride: 93.8 ± 4.6% of baseline, n = 5).

Figure 5.

A single in vivo exposure to cocaine impairs the induction of 5-hydroxytryptamine (5-HT)-long-term depression (LTD) via the activation of DA D1-like receptors. (a) Summary of experiments showing the induction of CP 93129-LTD in the nucleus accumbens (NAc) in slices from mice receiving SCH23390 (0.5 mg/kg, intraperitoneal) 15 min before saline (10 ml/kg, intraperitoneal; □) or cocaine (15 mg/kg, intraperitoneal; ■) injection. (b) Summary of experiments showing the induction of CP 93129-LTD in the NAc in slices from mice receiving raclopride (0.5 mg/kg, intraperitoneal) 15 min before saline (♢) or cocaine injection (♦). (c) Summary of experiments showing that bath application of CP 93129 (2 μM) for 10 min elicited a normal LTD in slice from mice that had received a single citalopram injection (10 mg/kg, intraperitoneal) 24 h earlier.

Because cocaine also has high binding affinity to the 5-HT transporters (Ritz et al. 1987), we explored the possibility that the loss of 5-HT-LTD observed in cocaine-treated mice may be because of increased levels of extracellular 5-HT in the NAc. To test the involvement of a serotonergic mechanism in cocaine's action, the effect of a 5-HT transporter inhibitor citalopram [10 mg/kg, intraperitoneal, a dose that has been shown to be effective to mimic the effect of cocaine (20 mg/kg, intraperitoneal) to block 5-HT transporter activity (Mateo et al. 2004)] on the induction of CP 93129-LTD was evaluated. In contrast to what was observed with CP 93129-LTD in cocaine-treated mice, no change in the extent of CP 93129-LTD was observed in slices from citalopram-treated mice (25–30 min after CP 93129 washout: 72.4 ± 5.8% of baseline, n = 4) compared with those from saline-treated mice (Fig. 5c). These findings indicate that the activation of DA D1-like receptors is responsible for the inhibitory effect of single cocaine exposure on 5-HT-LTD in the NAc.

A single in vivo exposure to cocaine impairs 5-HT-LTD via cAMP-dependent protein kinase signaling pathway

The preceding results point to an involvement of 5-HT1B receptors in the induction of 5-HT-LTD. We reasoned that the loss of 5-HT-LTD observed in cocaine-treated mice could be as a result of a downregulation of 5-HT1B receptors in the NAc. To test this possibility, we compared the expression level of 5-HT1B receptor protein in the NAc of saline- and cocaine-treated mice. As shown in Fig. 6a, no significant change in 5-HT1B receptor protein levels in the NAc was detected in cocaine-treated mice compared with saline-treated mice.

Figure 6.

A single in vivo exposure to cocaine increases phosphorylation of rabphilin 3A and synapsin 1 in the nucleus accumbens (NAc). (a) Representative immunoblot and summary histogram showing the levels of 5-HT1B receptors in the NAc slices from mice that had received a single saline (SAL; 1 ml/kg, intraperitoneal) or cocaine injection (COC; 15 mg/kg, intraperitoneal) 24 h earlier. (b) Representative immunoblot and summary histogram showing the levels of rabphilin 3A (Rab 3A) protein phosphorylated at serine 234 in the NAc slices from SAL- or COC-treated mice. (c) Representative immunoblot and summary histogram showing the levels of synapsin 1 (Syn 1) protein phosphorylated at serine 9 in the NAc slices from SAL- or COC-treated mice. (d) Representative immunoblot and summary histogram showing the effect of CP 93129 (2 μM) treatment for 10 min on the levels of Rab 3A protein phosphorylated at serine 234 in the NAc slices from SAL-, COC-, or SCH23390 (SCH) + COC-treated mice. (e) Representative immunoblot and summary histogram showing the effect of CP 93129 (2 μM) treatment for 10 min on the levels of Syn 1 protein phosphorylated at serine 9 in the NAc slices from SAL-, COC-, or SCH23390 (SCH) + COC-treated mice. The total number of animals examined is indicated by n in parenthesis. *< 0.05 compared with saline-treated controls.

Because DA D1-like receptors are coupled to the activation of adenylyl cyclase (Gingrich and Caron 1993) and 5-HT-LTD appears to involve a decrease in cAMP levels (Mathur et al. 2011), we next explored the possibility that cocaine-induced upregulation of cAMP-dependent protein kinase (PKA) activity could contribute to the impairment of 5-HT-LTD induction. As an initial test of this hypothesis, we assessed the phosphorylated and total levels of rabphilin 3A and synapsin 1, two well characterized substrates of PKA involved in the presynaptic regulation of neurotransmitter release (Lonart and Südhof 2001), in the NAc of saline- and cocaine-treated mice. To detect phosphorylation, we used phosphospecific antibodies raised against the PKA phosphorylation sites in these proteins, Ser234 for rabphilin 3A and Ser9 for synapsin 1 (Czernik et al. 1987; Lonart and Südhof 2001). The amounts of phosphorylated rabphilin 3A (Fig. 6b) and synapsin 1 (Fig. 6c) were significantly increased 24 h after a single in vivo exposure to cocaine. There were no differences in the total protein levels of rabphilin 3A and synapsin 1 in the NAc between saline- and cocaine-treated mice. More importantly, CP 93129-induced decreases in rabphilin 3A and synapsin 1 phosphorylation were significantly attenuated by cocaine treatment, which was prevented by coadministration of SCH23390 with cocaine (Fig. 6d and e).

If the impairment of 5-HT-LTD observed in cocaine-treated mice was associated with an increase in PKA activity, we would predict that the proportion of PKA contributing to the regulation of basal synaptic transmission might be increased. To test this hypothesis, we applied KT5720 (3 μM), an inhibitor of PKA, to monitor basal synaptic responses. Application of KT5720 for 40 min had no consistent effect on EPSCs in slices from saline-treated mice (96.2 ± 3.2% of baseline, n = 6; = 0.68; representative experiment in Fig. 7a and averaged data in Fig. 7c). In contrast, KT5720 caused a significant depression of EPSCs in slices from cocaine-treated mice (70.7 ± 6.8% of baseline, n = 8; < 0.01; representative experiment in Fig. 7b and averaged data in Fig. 7c).

Figure 7.

A single in vivo exposure to cocaine enhances KT5720-induced synaptic depression in the nucleus accumbens (NAc). (a) and (b) Example of an experiment showing the effects of protein kinase inhibitor KT5720 (3 μM) on excitatory postsynaptic currents (EPSCs) in slices from mice that had received a single saline (10 ml/kg, intraperitoneal) (a) or cocaine injection (15 mg/kg, intraperitoneal) (b) 24 h earlier. (c) Summary of experiments showing the effects of KT5720 (3 μM) treatment on the basal synaptic transmission is slices from saline (○) or cocaine-treated mice (●).

A single in vivo exposure to cocaine decreases the contribution of 5-HT1B receptors to subsequent cocaine-induced hyperlocomotion

Having confirmed the influence of a single cocaine exposure on 5-HT1B receptor function in the ex vivo slice preparations, we finally assessed the possible functional relevance of a loss of NAc 5-HT-LTD in cocaine-treated mice. Because 5-HT1B receptors in the NAc have been shown to play a permissive role in cocaine-induced hyperlocomotion (Neumaier et al. 2002; Przegaliński et al. 2002), we therefore examined whether a single cocaine administration may attenuate the contribution of 5-HT1B receptors to subsequent cocaine-induced hyperlocomotion. To test the involvement of 5-HT1B receptors in cocaine actions, the effect of 5-HT1B receptor antagonist NAS-181 (1 mg/kg, intraperitoneal) on cocaine (15 mg/kg, intraperitoneal)-induced hyperlocomotion was evaluated. Administration of NAS-181 alone had no significant effect basal locomotor activity, but it significantly attenuated the locomotor activation induced by cocaine (< 0.01 compared with cocaine alone; Fig. 8b). Using a 2-day protocol (Fig. 8a), a single cocaine (15 mg/kg) injection on day 1 abrogated the inhibitory effect of NAS-181 on cocaine-induced hyperlocomotion on day 2 (= 0.39 compared with cocaine alone; Fig. 8b), whereas the extent of total locomotor activity at 20 min intervals following cocaine injection was not altered by cocaine challenge 24 h earlier.

Figure 8.

A single in vivo exposure to cocaine decreases the contribution of 5-hydroxytryptamine (5-HT)1B receptors to cocaine-induced locomotor activation. (a) Schematic representation of the experimental designs for examining the contribution of 5-HT1B receptors in cocaine-induced hyperlocomotion. Following intraperitoneal injection of saline or cocaine, mice were immediately placed in the activity chamber and horizontal locomotor activity was monitored for 20 min. (b) Summary of experiments showing that cocaine (15 mg/kg, intraperitoneal)-induced locomotor activation was partially prevented by NAS-191 (1 mg/kg, intraperitoneal) pretreatment, which was abrogated by a single cocaine injection 24 h earlier. The total number of animals examined is indicated by n in parenthesis. *< 0.05 compared with saline-treated controls.

Discussion

This study provides further evidence supporting a role of the 5-HT system in the long-term regulation of glutamatergic transmission onto MSNs of the NAc. Our results indicate that pharmacological activation or synaptic stimulation of 5-HT1B receptors results in an induction of 5-HT-LTD and this LTD is mediated presynaptically. More importantly, we show that a single in vivo exposure to cocaine disables the ability of 5-HT to induce LTD. The molecular mechanisms mediating this dysregulation may involve an elevation of PKA activity subsequent to the rise in cAMP following DA D1-like receptor activation by cocaine treatment, which leads to increased phosphorylation of presynaptic proteins involved in synaptic vesicle release and subsequently counteracts the inhibitory effect of 5-HT1B receptors on glutamate release (Figure S1). To our knowledge, this is the first study to correlate a single in vivo exposure to cocaine with an alteration in 5-HT-LTD of glutamatergic transmission in the NAc. We also show that a single cocaine administration abrogates the contribution of 5-HT1B receptors to cocaine-induced hyperlocomotion.

5-HT exerts both presynaptic and postsynaptic neuromodulatory actions in several brain regions. A series of studies by Aghajanian and colleagues (Aghajanian and Marek 1999; Lambe et al. 2000) have revealed that 5-HT can enhance glutamatergic transmission in layer V pyramidal neurons of the frontal cortex through the activation of 5-HT2A receptors. It has also been reported that amphetamine substantially depresses glutamatergic transmission onto VTA neurons in the midbrain slice preparations via 5-HT receptor activation (Jones and Kauer 1999). The NAc receives rich serotonergic afferents from the midbrain dorsal raphe nucleus (Steinbusch 1981; Phelix and Broderick 1995; Van Bockstaele et al. 1996). Previous studies addressing the role of the NAc 5-HT system focused mainly on its contribution to the psychostimulant effects of cocaine (Müller and Huston 2006; Carey et al. 2008). Here, we have extended these observations by demonstrating that pharmacological activation or synaptic stimulation of 5-HT1B receptors can reliably induce LTD of glutamatergic transmission onto NAc MSNs, as they do in the dorsal striatum (Mathur et al. 2011). Three lines of evidence suggest that 5-HT-LTD is presynaptically mediated. First, CP 93129-LTD was associated with equal changes in AMPA and NMDA receptor-mediated components of EPSCs (Fig. 2a and b). Second, the expression of CP 93129-LTD was accompanied by an increase in the magnitude of PPR (Fig. 2c), which is generally considered to indicate a presynaptic mode of action (Zucker 1989). Third, CP 93129 significantly decreased the frequency of mEPSCs without affecting the mean current amplitude (Fig. 2d and e). A change in the amplitude of mEPSCs has traditionally been interpreted as a postsynaptic modification, whereas a decrease in their frequency is typically associated with mechanisms for inhibition of transmitter release. Furthermore, the lack of effect of CP 93129 on the amplitude of mEPSCs also implies that CP 93129-LTD is not mediated by a change in postsynaptic sensitivity to glutamate. These observations are in line with earlier reports indicating that 5-HT1B receptors are localized predominantly on axon terminals in the rodent brain and function as presynaptic heteroreceptors to regulate glutamate release (Bruinvels et al. 1994; Sari et al. 1997, 1999). Interestingly, Muramatsu et al. (1998) have reported, in contrast to our results, that application of 5-HT in the NAc only produced a reversible inhibition of glutamatergic transmission. This inconsistency may be a result of differences between studies in the duration and concentration of 5-HT application. In Muramatsu et al.'s (1998) study, 5-HT (10–30 μM) was bath applied to rat NAc slices for 5 min. However, in this study, 5-HT (5 μM) was applied to mouse NAc slices for 10 min. It is noteworthy that our results are consistent with a recent report showing that bath applying CP 93129 for 10 min resulted in LTD of glutamatergic transmission in the NAC core of rat slices (Mathur et al. 2011).

An important finding of this study is that a single in vivo exposure to cocaine impairs the induction of 5-HT-LTD in the NAc. Our finding seems to contradict the current model about the hierarchical link of cocaine-evoked synaptic plasticity between the VTA and the NAc (Mameli et al. 2009; Lüscher and Malenka 2011). For example, previous studies have demonstrated that a single cocaine injection is sufficient to cause a rapid but transient potentiation of glutamatergic inputs onto VTA DA neurons (Ungless et al. 2001; Argilli et al. 2008), whereas repeated cocaine injections are required to induce plasticity in the NAc (Thomas et al. 2001; Fourgeaud et al. 2004; Mameli et al. 2009). On the other hand, cocaine-evoked synaptic plasticity in the NAc occurs on a slower and longer lasting time course than the VTA. However, two previous studies found that a single in vivo exposure to either Δ9-tetrahydrocannabinol or cocaine is sufficient to abolish endocannabinoid-mediated LTD (eCB-LTD) in the NAc (Fourgeaud et al. 2004; Mato et al. 2005). In support of these findings, we have shown that the loss of NAc 5-HT-LTD observed in cocaine-treated mice was evident as early as 12 h after treatment and continued through 1 and 3 days. We also confirm that a single cocaine exposure caused no change in the ratio of AMPA receptor-mediated EPSCs to NMDA receptor-mediated EPSCs recorded from NAc core MSNs when measured 24 h later in brain slices (data not shown). The most straightforward interpretation of these observations is that a single cocaine exposure may induce fast-acting adaptive responses to alter the inducibility of some forms of synaptic plasticity in the NAc (e.g., eCB-LTD and 5-HT-LTD), whereas repeated cocaine exposure and withdrawal may lead to prolonged modifications of synaptic weights that generally inhibit further induction of synaptic plasticity.

Our results also demonstrate that cocaine-induced impairment of 5-HT-LTD in the NAc is dependent on DA D1-like receptor activation, since the effect of cocaine was prevented when cocaine was coadministered with the selective DA D1 receptor antagonist SCH23390. Given that the 5-HT system has also been shown to play a vital role in various effects of cocaine in the brain (Müller and Huston 2006), it might be argued that cocaine-induced impairment of 5-HT-LTD seen in this study is in part through its direct action on the 5-HT system. However, this possibility appears unlikely because pharmacological blockade of 5-HT transporter activity with citalopram cannot mimic the effect of cocaine on 5-HT-LTD. Although these experiments do not allow us to determine where in the brain DA D1-like receptors are required to trigger the inhibitory effect of cocaine on 5-HT-LTD, our findings are in line with an earlier report indicating that a single cocaine administration abolishes eCB-LTD in the NAc through a DA D1 receptor-mediated mechanism (Fourgeaud et al. 2004). Our finding that SCH23390 prevented the inhibitory effects of cocaine on CP93129-induced decreases in presynaptic rabphilin 3A and synapsin 1 phosphorylation (Fig. 6d and e), supports the location of DA D1 receptors on the presynaptic compartment of glutamatergic terminals. However, further studies are required to examine this possibility.

We propose that cocaine-induced upregulation of PKA activity in the NAc could contribute, at least in part, to the impairment of 5-HT-LTD, as evidenced by increased phosphorylation of two presynaptic PKA substrates, rabphilin 3A and synapsin 1 (Fig. 6b and c), and by attenuating the ability of CP 93129 to decrease rabphilin 3A and synapsin 1 phosphorylation (Fig. 6d and e). Furthermore, our data confirmed a marked increase in KT5720-mediated depression of basal synaptic transmission in NAc slices from cocaine-treated mice (Fig. 7). Although the precise mechanism by which PKA inhibits 5-HT1B receptor function remains unclear, it seems plausible that the elevations of PKA activity may lead to increased phosphorylation of presynaptic proteins involved in synaptic vesicle release machinery and thereby counteract the inhibitory effect of 5-HT1B receptors on glutamate release. Indeed, previous work has reported that forskolin or membrane-permeable cAMP analog 8-Br-cAMP significantly block CP 93129-LTD at corticostriatal synapses, suggesting that 5-HT1B receptors work through the cAMP-dependent pathway to induce LTD (Mathur et al. 2011).

What could be the functional relevance of cocaine-induced impairment of 5-HT-LTD in the NAc? Considering that 5-HT1B receptors in the NAc have been reported to play a facilitatory role in cocaine-induced hyperlocomotion and behavioral sensitization (Neumaier et al. 2002; Przegaliński et al. 2002), this loss of 5-HT1B receptor function would be expected to decrease the behavioral responses to cocaine. This study also provides novel evidence in supporting the view that a single cocaine exposure is associated with decreased 5-HT1B receptor function, which may provide a mechanism to fine-tune the effects of 5-HT on glutamatergic plasticity in the NAc. Given that an LTD-like process in the NAc has been implicated to contribute to the reorganization of neural circuitry that underlies behavioral sensitization to cocaine (Thomas et al. 2001), we therefore speculate that the loss of ability to undergo LTD in the NAc of cocaine-exposed mice might serve as a compensatory mechanism to counterbalance decreased glutamatergic activity measured in response to cocaine (Fourgeaud et al. 2004). In line with this hypothesis, it has been demonstrated that a single in vivo cocaine administration abolished endocannabinoid-mediated LTD in the NAc (Fourgeaud et al. 2004).

In conclusion, the results from this study show that a single exposure to cocaine can impair 5-HT-LTD in the NAc by a PKA-dependent mechanism. Our data also indicate that higher levels of PKA activity subsequent to the rise in cAMP following DA D1-like receptor activation may promote the neurotransmitter release machinery and, consequently, serve as a compensatory mechanism to counteract the inhibitory effect of 5-HT1B receptors on glutamate release. These findings provide major advances in our understanding of the functional role of the 5-HT system in the NAc.

Acknowledgements

This study was supported by research grants (NSC99-2321-B-006-012-MY3; principal investigator, C.C.H. and NSC100-2321-B-006-001; principal investigator, K.S.H.) from the National Science Council, Taiwan.

Conflict of interest

The authors declare no conflict of interest.

Ancillary