Abstract

In rat prefrontal cortex (PFC), long-term depression induced by low-frequency single stimuli has never been studied. Combined with the well-documented involvement of dopamine transporters (DATs) in the regulation of PFC-dependent cognitive processes, it is important to test whether this form of plasticity can be modulated by DAT activity in the PFC. Here, we show first that prolonged 3-Hz stimuli successfully induced synaptic depression in rat PFC slices whose induction depended on endogenous stimulation of D1-like and D2-like receptors and the activation of extracellular signal-regulated kinase 1/2 (ERK1/2). This depression was found to be significantly impaired by selective inhibition of the DAT by GBR12909 (1–200 nM) or GBR12935 (100 nM). The excess amount of extracellular dopamine caused by DAT inhibition acted critically on D1-like receptors to impair depression. Furthermore, this impairment by GBR12 909 was cancelled by the allosteric-positive mGluR5 modulator CDPPB, the drug known to reverse hyperdopaminergia-induced abnormal PFC activity, and the associated cognitive disturbances. Finally, these induction, impairment, and restoration of synaptic depression were correlated by an inverted-U shape manner with the phosphorylation level of ERK1/2. We suggest that abnormal increases of the extracellular dopamine level by DAT inhibition impair synaptic depression in the PFC through over-stimulation of D1-like receptors.

Introduction

Long-term potentiation (LTP) and long-term depression (LTD) are well-studied memory models in brain areas such as hippocampus (Malenka and Bear 2004; Pastalkova et al. 2006; Nicholls et al. 2008). In the prefrontal cortex (PFC), however, it is still unclear how LTP and LTD play roles for the function of this cognitively important brain area. This appears to be partly because it was long thought that the PFC executive function (Fuster 1995) is independent of lasting plastic changes of PFC neurons. But there seems to be a consensus now that neurons in the PFC, as well as in the hippocampus, store traces for the cognitive aspect of temporal organization of behavior (Goto et al. 2010; DeVito and Eichenbaum 2011). We suggested that this kind of long-term memory, that is, abstract memory for behavioral structure or rule, might be supported by long-term synaptic plasticity in the PFC (Goto et al. 2010).

It is thought that synaptic plasticity in the PFC can take aberrant forms (Matsuda et al. 2006; Kolomiets et al. 2009). Aberrant plasticity in the PFC may underlie cognitive deficits seen in certain psychiatric disorders through interfering with physiological neural traces in the PFC (Goto et al. 2010). For example, the cognitive symptoms in schizophrenia include a disability to form goal-directed action plans (Barch and Dowd 2010). Such a deficit may be at least in part related to disrupted inductions of neuroplasticity in the PFC. Supporting this view, the patients of schizophrenia indeed exhibit dysfunctional plasticity in their brain (Frantseva et al. 2008).

Dopamine is a powerful modulator of synaptic plasticity in the PFC, exerting its effect in a so-called “inverted-U” shape dose-dependent manner (Goto et al. 2010). Thus, a novel form of aberrant neuroplasticity in the PFC may be detected under the blockade of dopamine transporters (DATs), which regulate the extracellular level of dopamine in the brain (Giros and Caron 1993). Indeed, a modified activity of the DAT has been implicated in a number of psychiatric disorders that may involve PFC dysfunction, including drug addiction (Reith et al. 1997), attention-deficit/hyperactive disorder (ADHD; Gainetdinov and Caron 2001), depression (Perona et al. 2008), and bipolar disorder (Greenwood et al. 2006). In addition to these disorders, furthermore, inhibition of the DAT, particularly in the PFC, is thought to be the main cause for the reinstatement of cocaine-induced conditioned place preference (CPP) in rodents (Sanchez et al. 2003; Schmidt and Pierce 2006), that is, a rigid form of goal-direction memory likely to derive from the lack of flexible control of behavior (Jentsch et al. 2002).

Based on the involvement of DAT activity in the aforementioned many dysfunctional cognitive states, it appears important to test whether inhibition of the DAT affects synaptic plasticity in the PFC. In this context, it is particularly relevant to monitor synaptic depression induced by low-frequency single stimuli, since this form of plasticity is believed to serve as a cellular basis for behavioral flexibility (Morice et al. 2007; Nicholls et al. 2008), that is, the cognitive domain disrupted in the cocaine-induced CPP mentioned above. Thus, we hypothesize that the inhibition of the DAT may block this form of synaptic depression in the PFC. Furthermore, low-frequency, single stimuli-induced LTD has never been studied in the rat PFC. In the present study, we therefore tested whether selective pharmacological inhibition of the DAT impairs synaptic depression induced by low-frequency single stimuli.

Materials and Methods

Slice Preparation

All experiments were conducted in accordance with Regional Committee for the ethics of animal experimentation. Male Sprague-Dawley rats (25–30-day-old juvenile rats or 6–7-week–old adult rats) were decapitated, and their brains were rapidly removed from their skull. We used no anesthetics upon the sacrifice to avoid possible intervening effects of anesthetic agents on synaptic transmission and plasticity, but thorough care was taken to minimize the suffering of the animals. Coronal slices containing the prelimbic area (300–400 μm; 2.2–3.7 mm from the bregma) were sectioned by the use of a Campden vibratome (Campden Instruments, UK) in chilled (≍0–4°C) oxygenated (95% O2/5% CO2) artificial cerebrospinal fluid (ACSF) of the following composition (mM): NaCl 124, KCl 2, NaHCO3 26, KH2PO4 1.15, MgCl2 1, CaCl2 2, and d-glucose 11. The slices were allowed to recover for at least 3 h at room temperature (≍20°C) in oxygenated ACSF. Experiments were performed in a submersion-recording chamber with continuous perfusion (1 mL/min) of warmed ACSF (32°C).

Stimulation and Recording

Field Potential Recording

The protocol for field potential recording followed that of Morris et al. (1999). Upon the layer I–II single stimulation (monophasic constant current square pulses; 100 μs width; A360 stimulus isolator, WPI, FL) through a bipolar stimulating electrode (Teflon-coated tungsten wire, external diameter 125 μm; A-M Systems, WA) placed on slice surface, we could detect a series of negative- and positive-going field responses along the depth of the pial surface–white matter axis in the prelimbic area (Fig. 1A1; with glass electrodes filled with 3 M NaCl), consistent with Morris et al. (1999). The sharp, negative-going postsynaptic potential recorded at layers I–II is likely to contain dendritically evoked synaptic responses of layer III and V pyramidal neurons, where the layer V pyramidal neurons are thought to be the major source (Morris et al. 1999). This potential was far more robust and stable than responses recorded at deeper layers. Our pharmacological characterization clearly indicated that this response is glutamatergic excitatory postsynaptic potential (EPSP), since it was abolished by bath application of α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA)-receptor antagonist 6-cyano-7-nitroquinoxaline-2,3-dion (CNQX, 10 μM) and N-methyl-d-aspartate (NMDA)-receptor antagonist DL-2-amino-5-phosphonopentanoic acid (DL-AP5, 100 μM), in the presence of γ-aminobutyric acid (GABA)-A-receptor antagonist bicuculline (2 μM; Fig. 1A2). We therefore used this layer I–II potential in our experiments. The response size was set at 70% of maximum amplitude, which is usually about 0.7 mV with 20- to 40-μA stimulus intensity.

Figure 1.

Low-frequency stimuli-induced LTD in rat PFC slices. (A1) Depth profile of the field response evoked by layer I–II stimulation, following Morris et al. (1999). (A2) The layer I–II response was abolished by bath application of CNQX (10 μM) and DL-AP5 (100 μM) (in the presence of 2 μM bicuculline). (B) Three-Hz stimuli (15 min) induced LTD of the layer I–II response. Representative mean responses taken before and 45 min after 3 Hz stimuli are shown. (C) This LTD was expressed also as a downward shift of the input–output curve taken after 3-Hz stimuli (IO2) compared with that taken before stimuli (IO1). (D) LTD was also recorded intracellularly from single layer V pyramidal neurons. There were no signs of decline in membrane potential (inset), input resistance, and the spike height during the course of the experiments (see text). Representative mean responses taken before and 45 min after 3-Hz stimuli are also shown.

Figure 1.

Low-frequency stimuli-induced LTD in rat PFC slices. (A1) Depth profile of the field response evoked by layer I–II stimulation, following Morris et al. (1999). (A2) The layer I–II response was abolished by bath application of CNQX (10 μM) and DL-AP5 (100 μM) (in the presence of 2 μM bicuculline). (B) Three-Hz stimuli (15 min) induced LTD of the layer I–II response. Representative mean responses taken before and 45 min after 3 Hz stimuli are shown. (C) This LTD was expressed also as a downward shift of the input–output curve taken after 3-Hz stimuli (IO2) compared with that taken before stimuli (IO1). (D) LTD was also recorded intracellularly from single layer V pyramidal neurons. There were no signs of decline in membrane potential (inset), input resistance, and the spike height during the course of the experiments (see text). Representative mean responses taken before and 45 min after 3-Hz stimuli are also shown.

Intracellular Recording

The soma of layer V pyramidal neurons was penetrated with a sharp glass micropipette (GC120F-10, Harvard Apparatus, Holliston, MA) filled with 3 M K-acetate (80–120 MΩ tip resistance), which is our routine preparation (Kolomiets et al. 2009; Marzo et al. 2010). Initially, negative currents were injected with an Axoclamp 2A (Molecular Devices, Union City, CA) or a BVC-700A amplifier (Dagan Corporation, MI) to help stabilize the cells. After their stabilization, all or most currents were removed. As a practice, the membrane was held more negative than resting membrane potential when the difference between spike threshold (−54.0 ± 1.7 mV on average as verified postrecording) and the membrane potential was found smaller than 15 mV. This procedure was practiced in order to evoke sufficiently large monosynaptic EPSPs for good off-line quantification. The average resting membrane potential as verified by postexperiments was −71.6 ± 1.7 mV, and the potential held during experiments was −74.1 ± 0.9 mV. In addition, the initial and final spike heights were 77.8 ± 2.8 and 74.4 ± 2.6 mV, respectively. The initial and final input resistances were 51.7 ± 7.5 and 54.4 ± 4.8 MΩ, respectively. Only cells that remained within 10% of the initial values of these parameters were included for later analysis. For response recording, EPSP of approximately 10 mV in amplitude was evoked by single-pulse stimulation to layer I–II fibers.

Experimental Procedures

In both extracellular and intracellular configurations, all experiments were in principle conducted in the following order: the recording of baseline EPSPs for at least 20 min, 10–15 min application of receptor antagonists or kinase inhibitors in the case of pharmacological experiments, and the conditioning stimuli (3-Hz repetitive stimulation for 15 min) that induce plasticity. Drugs were washed immediately after the end of the 3-Hz stimulation. The responses were followed for 50 min postconditioning.

In the principal groups, just before starting experiments and at the end of the experiments, input–output relation was tested. The response was evoked at 0.1 Hz to ascending stimulus intensities ranging from 10 to 100 μA delivered by a 10-μA step. Three responses were collected at each of these stimulus intensities.

In the majority of cases, experiments were performed in the presence of GABA-A antagonist bicuculline methiodide (1 μM) in bathing medium. The use of bicuculline has been often practiced in plasticity research to reduce inhibition and to isolate excitatory transmission (e.g. Wigström and Gustafsson 1983; Artola and Singer 1987; Hirsch and Crepel 1990). However, we conducted our principal LTD experiment also in the absence of bicuculline. We found that LTD is clearly inducible under intact synaptic inhibition (see Supplementary Fig. SI2).

In a given set of pharmacological condition, the experiments were performed in the interleaved manner. All drugs were included in perfusion medium. All evoked responses were fed to an amplifier in the current-clamp mode and digitized at 10 kHz through a Digidata 1322A interface (Molecular Devices, CA) by the use of Elphy data acquisition–analysis program developed by Dr G. Sadoc (Institut Alfred Fessard, CNRS, Gif-sur-Yvette, France).

Analysis

For analysis, the initial slope of EPSP (≤1 ms period from its onset, mV/ms) was calculated for each individual EPSP. The initial slope measurement has been routinely practiced in different studies including our owns (Hirsch and Crepel 1990; Huang et al. 2004; Kolomiets et al. 2009; Marzo et al. 2010), since it contains only monosynaptic component of the responses (Hirsch and Crepel 1990) and since the measurement of response amplitude is contaminated by the successively occurring polysynaptic responses. Changes of the EPSP slope after conditioning stimuli (40–50 min after conditioning; denoted as “45 min after conditioning”) were then expressed as a percent increase or decrease from its preconditioning baseline level (the 10-min period just before the conditioning stimuli or drug application). In the case of intracellular recording, in order to reduce variability, the percentage values were grouped for each successive 2-min period (i.e. 4 responses). These percentage values were then compared between different experimental groups for statistics. A repeated-measure analysis of variance (ANOVA) was used to compare increases/decreases seen during the entire postconditioning period. The 2-tailed Student t-test was used to compare EPSP changes seen at 45 min after conditioning. In some cases, mean raw EPSP slope values (mV/ms) calculated from 10-min period just before 3-Hz stimulation (baseline) and from 40–50 min after stimulation (LTD) were compared within a group by paired t-test to signify LTD induction in that group. Equally, to control the action of inhibitory drugs on basal synaptic transmission, in separate slices, we applied drugs for the period equivalent to the relevant preconditioning perfusion period plus 3-Hz conditioning stimuli delivery (15 min) and followed the responses for another 50 min. The predrug baseline and the time period equivalent to “45 min after conditioning” were compared by paired t-test to ensure the stability of EPSP in the presence of the drugs.

The drugs used were: bicuculline methiodide (Sigma Aldrich, France), CNQX (Tocris, UK), DL-AP5 (Tocris), dopamine (Sigma Aldrich), SCH23390 (Tocris), sulpiride (Tocris), and PD98059 (Calbiochem, Germany), GBR12909 (Sigma Aldrich), GBR12935 (Sigma Aldrich), RO31-8220 (Tocris), U73122 (Calbiochem).

Western Blot

At the end of the experiments, the brain slices were placed on ice, and the area corresponding to the prelimbic area was rapidly dissected and transferred to ice cold lysis buffer (10 mM Tris–HCl, 50 mM NaCl, 1% Triton X-100, 30 mM sodium pyrophosphate, 50 mM NaF, 5 μM ZnCl2, and 0.5 mM DTT) and a mix of protein protease and phosphatase inhibitors (100 mM Na3VO4, 100 nM okadaic acid, 2.5 μg/mL aprotinin, 2.5 μg/mL pepstatin, 0.5 mM phenylmethylsulfonyl fluoride, 0.5 mM benzamidine, and 2.5 μg/mL leupeptin). Tissues were then homogenized and centrifuged (13 000 × g; 20 min; 4°C) to pellet insoluble materials. Lysates (10–30 μg) were boiled for 5 min and analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis on 10% gels. After blocking nonspecific sites with 5% nonfat dry milk, blots were incubated overnight at 4°C with antiphospho-(Thr202–Tyr204)-extracellular signal-regulated kinase 1/2 (ERK1/2) primary antibody (1:5000; Cell Signalling Technology, Saint Quentin en Yvelines, France) or anti-ERK antibody (1:5000; Santa Cruz Biotechnology, Tebu, Le Perray-en-Yvelines, France) in the blocking solution with 0.1% Tween-20. Blots were then extensively washed before being incubated for 2 h at room temperature with antirabbit or antigoat horseradish peroxidase (HRP)-conjugated antibodies (Amersham Pharmacia Biotech, Sweden) diluted to 1:5000 in the blocking solution with 0.1% Tween-20. Proteins were detected by chemiluminescence using the ECL kit (GE Healthcare, Saclay, France). As a loading control, blots were subsequently incubated with a β-tubulin antibody (1:5000, Sigma Aldrich) and an antimouse HRP-conjugated secondary antibody. Films were scanned and analyzed using the Scion Image software. Relative ERK expression or phosphorylation levels were measured by normalization of the optical density obtained from either the total ERK or the phospho-ERK signal with that of β-tubulin.

Results

Low-Frequency Stimulation Induces Synaptic Depression in Rat PFC Neurons

The principal experiments were conducted in PFC slices prepared from 25–30-day-old male Sprague-Dawley rats in the presence of 1-μM GABA-A-receptor antagonist bicuculline (but see below and Supplementary Fig. SI2 for other conditions). The protocol for the field potential recording was previously established by Morris et al. (1999) and adopted in the present study (Fig. 1A1). The large and stable negative-going potential recorded in layers I–II upon stimulation to layer I–II afferent fibers is the glutamatergic synaptic response, as it was abolished by bath application of AMPA-receptor antagonist CNQX (10 μM) and NMDA-receptor antagonist DL-AP5 (100 μM; Fig. 1A2; see Materials and Methods for details of the depth profile of responses). After a stable baseline recording for at least 20 min, low-frequency stimuli (3 Hz for 15 min) were delivered to layer I–II fibers. This stimulus protocol was chosen as it had been shown to induce LTD in the mouse PFC (Huang et al. 2004) and as our pilot experiments suggested that the standard 1-Hz stimuli (Malenka and Bear 2004) or paired-pulse stimuli delivered at a low-frequency (Otani and Connor 1995) did not reliably induce LTD in the rat PFC.

As shown in Figure 1B, the 3-Hz single stimuli induced a stable depression of the EPSP slope (−23.7 ± 6.4% decrease 45 min after conditioning stimuli, n = 7, P < 0.025, paired t-test, Fig. 1B). This depression was also expressed as a downward shift of the input–output curve collected just before and after the experiments (see Materials and Methods for input–output curve collection; Fig. 1C). Repeated-measure 2-way ANOVA indicated a significant main effect (F1,14 = 16.448, P < 0.002) and a significant group × stimulus intensity interaction (F9,126 = 2.763, P < 0.01) between these 2 curves. Thus, the 3-Hz stimuli induced synaptic depression of PFC glutamatergic synaptic responses that lasted for 50 min (see “Analysis” of Materials and Methods). Hereafter, whenever appropriate, we refer to this synaptic depression as “early LTD” because a few hour maintenance is often taken as a standard characteristic for LTP/LTD under in vivo conditions, although approximately 1 h monitoring is frequently practiced in slice studies. We note, however, that the current synaptic depression was still well maintained 50 min after induction.

We verified that this early LTD did not result from nonphysiological declines of neuronal parameters by the use of intracellular recordings. The glutamatergic EPSP was recorded from the soma of layer V pyramidal neurons upon stimulation to layer I–II fibers (Kolomiets et al. 2009; Marzo et al. 2010). The response from layer V pyramidal neurons is thought to comprise the major source for the extracellularly recorded field response under the same configuration (Morris et al. 1999). Under this condition, 3-Hz stimuli successfully induced a depression of the EPSP that lasted at least for 50 min (−24.8 ± 8.0% 45 min after stimulation, n = 9, P < 0.02, paired t-test, Fig. 1D), accompanied by no signs of decline in the resting membrane potential (−74.6 ± 0.7 mV at the start of experiments vs. −74.1 ± 0.9 mV at the end of experiments, Fig. 1D), input resistance (51.7 ± 7.5 vs. 54.4 ± 4.8 MΩ), and the spike height (77.8 ± 2.8 vs. 74.4 ± 2.6 mV), indicating that the present depression is a genuine form of synaptic depression.

We further showed that this early LTD is NMDA receptor-dependent (see Supplementary Fig. SI1) and provided evidence for the physiological relevance of this LTD (see Supplementary Fig. SI2) by showing that 1) this LTD was still inducible in the absence of bicuculline, 2) this LTD was induced in slices prepared from adult rats (6–7 weeks), and 3) this LTD occurred in the presence of background dopamine unlike the previously reported LTD induced by high-frequency stimuli (Matsuda et al. 2006; Kolomiets et al. 2009), which we consider as a dysfunctional form of synaptic depression (Goto et al. 2010).

Stimuli-Induced LTD Depends on Endogenous Dopamine Acting on D1 and D2 Receptors

We next tested whether this early LTD depends on the activation of dopamine receptors. Dopamine is a critical neurotransmitter for PFC function (Zahrt et al. 1997; Sanchez et al. 2003; Williams and Castner 2006). Many forms of synaptic plasticity in the PFC are dopamine dependent (Gurden et al. 1999; Otani et al. 1999; Matsuda et al. 2006; Kolomiets et al. 2009). In PFC slices, dopamine is endogenously released from residual axon terminals upon repetitive stimuli and induces plasticity (Young and Yang 2005; Kolomiets et al. 2009). First, we bath-applied both D1-like receptor antagonist SCH23390 (2 μM; Huang et al. 2004) that blocks dopamine D1 and D5 receptors and D2-like antagonist sulpiride (20 μM; Gorelova et al. 2002) that blocks dopamine D2 and D3 receptors, together from 10 min before 3-Hz stimuli until the end of the stimuli (n = 6). Our separate control experiments verified that these antagonists do not significantly change baseline responses for the time period equivalent to preconditioning drug perfusion + conditioning stimuli delivery + plasticity monitoring (response change at the time equivalent to 45 min after 3-Hz stimulation in this group was −1.6 ± 7.1%, n = 8; P > 0.1 by paired t-test). Under this antagonist condition, synaptic depression initially appeared after 3-Hz stimuli, but it rapidly disappeared within approximately 20 min, resulting in a negligible change as measured 45 min after induction (−0.6 ± 3.9%, n = 6, Fig. 2A, filled diamond). There was a significant main effect between this group and the normal LTD group depicted in Figure 1B during the post 3-Hz stimuli period (F1,11 = 5.611, P < 0.05). The EPSP slope change at 45 min after 3 Hz stimuli in this group (−0.6 ± 3.9%) was significantly smaller than the normal LTD (−23.7 ± 6.4%, P < 0.01, 2-tailed t-test). These results show that this early LTD is triggered by endogenous dopamine acting on D1-like and D2-like receptors during 3-Hz stimuli.

Figure 2.

LTD induced by 3-Hz stimuli required concurrent endogenous stimulation of D1- and D2-like dopamine receptors. (A) Coapplication of D1-like receptor antagonist SCH23390 (1 μM) and D2-like receptor antagonist sulpiride (20 μM) blocked LTD (filled diamond). Representative mean responses taken before drug application and 45 min after 3-Hz stimuli are presented, showing the blockade of LTD. (B) Sole application of SCH23390 (1 μM) blocked LTD (filled triangle). (C) Sole application of sulpiride (20 μM) blocked LTD (filled circle).

Figure 2.

LTD induced by 3-Hz stimuli required concurrent endogenous stimulation of D1- and D2-like dopamine receptors. (A) Coapplication of D1-like receptor antagonist SCH23390 (1 μM) and D2-like receptor antagonist sulpiride (20 μM) blocked LTD (filled diamond). Representative mean responses taken before drug application and 45 min after 3-Hz stimuli are presented, showing the blockade of LTD. (B) Sole application of SCH23390 (1 μM) blocked LTD (filled triangle). (C) Sole application of sulpiride (20 μM) blocked LTD (filled circle).

To test which subtype of dopamine receptors is critically involved in this LTD, either SCH23390 (2 μM, n = 6) or sulpiride (20 μM, n = 6) was applied as in the above experiments. Our separate control experiments again verified that either of these antagonist applied alone does not significantly change the baseline level of responses during the time period equivalent to preconditioning drug perfusion + conditioning stimuli delivery + plasticity monitoring (changes at the time equivalent to 45 min after 3 Hz stimuli were: SCH23390 group, −3.8 ± 4.6%, n = 6; sulpiride group −1.1 ± 8.4%, n = 7; P > 0.1 by paired t-test). Under this condition, both antagonists significantly reduced the early LTD (−7.5 ± 2.5% and −3.4 ± 4.6% at 45 min, respectively; Fig. 2B, filled triangle; C, filled circle). ANOVA indicated that the difference during the post 3-Hz stimuli period between the SCH23390 and normal LTD groups shows a trend (F1,12 = 4.213, P = 0.0626) and that there was a highly significant group × time interaction (F99,1188 = 2.226, P < 0.0001). The EPSP change 45 min after 3-Hz stimuli in this SCH23390 group (−7.5 ± 2.5%) was also significantly smaller than the normal LTD (P < 0.03, 2-tailed t-test). Thus, SCH23390 particularly promoted the decay of LTD. In the sulpiride group, there were a significant main effect and a significant group × time interaction compared with the normal LTD group (F1,11 = 5.872, P < 0.05; F99,1089 = 1.833, P < 0.001, respectively). The EPSP change at 45 min after 3-Hz stimuli in this sulpiride group (−3.4 ± 4.6%) was also significantly smaller than the normal LTD (P < 0.03, by 2-tailed t-test). We concluded that a concurrent activation of D1-like and D2-like receptors is necessary for the full expression of early LTD where the role played by D2-like receptors appears larger than that by D1-like receptors.

Impairment of LTD by DAT Inhibition

Results so far indicate that the endogenous dopamine released upon 3-Hz stimuli triggers early LTD. We then tested whether a nonphysiological increase of dopaminergic tone by DAT inhibition alters this LTD. We used the specific DAT inhibitor GBR12909 that has inhibitory concentration 50 (IC50) value as low as a few nM for dopamine and has >400 times more specificity to dopamine reuptake than noradrenaline reuptake (Andersen 1989; Reith et al. 1994). It has been shown indeed that GBR12909 at a concentration as low as 10 nM significantly alters catecholamine release in prelimbic slices (∼300% increase in Km value: Mundorf et al. 2001). Accordingly, we bath-applied ranges of nM concentrations of GBR12909 (1–5 nM, n = 7; 50 nM, n = 7; and 200 nM, n = 5) from 15 min before 3-Hz stimuli until the end of the stimuli. We verified in separate slices that GBR12909 does not change baseline synaptic responses. Thus, as the representative concentrations, we applied either 5 nM (the frequently selected low concentration in the subsequent experiments) or 200 nM (the highest concentration used) for the duration of time equivalent to preconditioning drug perfusion + conditioning stimuli delivery + plasticity monitoring. GBR12909 did not significantly change baseline responses (at the time equivalent to 45 min after 3-Hz stimuli; 5 nM group, −2.1 ± 5.8%, n = 5; 200 nM group, −2.9 ± 3.7%, n = 6, P > 0.1 by paired t-test). Under this condition, GBR12909 significantly impaired LTD at each concentration range (Fig. 3A, filled triangle, filled diamond, and cross). There was a significant main effect during the post 3-Hz stimuli period between the normal LTD group and each of the GBR12909 groups (1–5 nM group F1,12 = 7.623, P < 0.02; 50 nM group F1,12 = 10.001, P < 0.01; 200 nM group F1,10 = 4.984, P < 0.05). EPSP slope changes 45 min after 3-Hz stimuli in these GBR groups were significantly different from the normal LTD (respectively, −1.7 ± 4.1%, 10.9 ± 8.6%, and −1.9 ± 5.1%, P < 0.01, P < 0.005, and P < 0.03, 2-tailed t-test, Fig. 3B). The input–output curve (1–5 nM group was taken as the representative) showed an overlap between the pre- and poststimulation periods (Fig. 3C), showing neither a main effect nor a group × stimulus intensity interaction by ANOVA analysis (F1,12 = 0.107, P > 0.5; F9,108 = 0.237, P > 0.9). To confirm this result, we used another DAT inhibitor GBR12935 (100 nM; Andersen 1989). GBR12935 also significantly reduced LTD (−10.3 ± 4.6% at 45 min, n = 10, P < 0.05 compared with the normal LTD by ANOVA [F1,15 = 4.913] and by 2-tailed t-test, Figure 3D, filled diamond). GBR12935 itself did not change baseline responses during the response monitoring for the equivalent time (−1.8 ± 4.2% at the time equivalent to 45 min after 3-Hz stimuli, n = 8). Thus, inhibition of DAT activity impaired 3-Hz stimuli-induced early LTD.

Figure 3.

Selective inhibition of DAT activity blocked 3-Hz stimuli-induced LTD. (A) Different concentrations of GBR12909 all blocked LTD (open triangle 1–5, filled diamond 50, and times 200 nM). For the 1–5-nM group, representative mean responses taken before GBR12909 application and 45 min after 3-Hz stimuli are also shown. (B) Summary graph of EPSP changes 45 min after 3-Hz stimuli in different conditions. “3 Hz” annotates the group where LTD was induced in the control solution. (C) Clear overlap was seen in input–output curves taken before and after experiments in the presence of GBR12909 (1–5 nM). (D) Another selective DAT inhibitor GBR12935 (100 nM) blocked LTD (filled diamond).

Figure 3.

Selective inhibition of DAT activity blocked 3-Hz stimuli-induced LTD. (A) Different concentrations of GBR12909 all blocked LTD (open triangle 1–5, filled diamond 50, and times 200 nM). For the 1–5-nM group, representative mean responses taken before GBR12909 application and 45 min after 3-Hz stimuli are also shown. (B) Summary graph of EPSP changes 45 min after 3-Hz stimuli in different conditions. “3 Hz” annotates the group where LTD was induced in the control solution. (C) Clear overlap was seen in input–output curves taken before and after experiments in the presence of GBR12909 (1–5 nM). (D) Another selective DAT inhibitor GBR12935 (100 nM) blocked LTD (filled diamond).

DAT Inhibitor Impairs LTD by Changing the Dopamine Level

The above results indicate that too low or too high dopaminergic tone impairs LTD, suggesting an “inverted-U” shape relation between this early LTD induction and the extracellular dopamine level. To confirm this result further, we tested the effect of low concentrations of dopamine antagonists on the GBR12909-induced LTD impairment. We reason that if too high levels of extracellular dopamine cause the LTD impairment, the dopamine antagonists at a low concentration should restore LTD by competing with the GBR12909 effect and reducing the level of dopamine-receptor stimulation. We thus coapplied SCH23390 (1 instead of 2 μM) and sulpiride (10 instead of 20 μM) from 5 min after GBR12909 (5 nM was used as the representative concentration) application until the end of the application/3-Hz stimuli. These antagonists (1 and 10 μM) themselves reduced LTD as tested in separate slices (−6.8 ± 5.9% at 45 min, n = 12; not significant from the baseline, P > 0.2 by paired t-test, see Supplementary Fig. S3 and figure legend). Under this condition, GBR12909-induced LTD impairment was cancelled so that a significant LTD appeared in the presence of the dopamine antagonists (−22.8 ± 5.4% 45 min after 3-Hz stimuli, Fig. 4A, filled triangle). There was a significant main effect between this group and the group where LTD was impaired by GBR12909 (1–5 nM; F1,11 = 11.789, P < 0.01). The EPSP slope change 45 min after 3-Hz stimuli in this group (−22.8 ± 5.4%) was significantly larger than that in the LTD impairment group (−1.7 ± 4.1%; P < 0.01). These results suggest that the GBR12909-induced LTD impairment indeed occurred through an increased dopaminergic tone.

Figure 4.

GBR12909-induced LTD block was caused by over-stimulation of D1-like receptors. (A) Low concentrations of SCH23390 (1 instead of 2 μM) and sulpiride (10 instead of 20 μM) were co-applied to counteract an increased dopamine level induced by GBR12909. LTD blockade by GBR12909 (Fig. 3A) was cancelled (filled triangle). Representative mean responses taken before the application of GBR12909 + SCH23390 + sulpiride and 45 min after 3-Hz stimuli are presented, showing the cancellation of the LTD blockade. (B) The same cancellation of GBR12909-induced LTD block was achieved by SCH23390 alone (filled diamond). (C) sulpiride alone failed to cancel the GBR12909-induced LTD block (open triangle) (see text for 20 μM sulpiride effect).

Figure 4.

GBR12909-induced LTD block was caused by over-stimulation of D1-like receptors. (A) Low concentrations of SCH23390 (1 instead of 2 μM) and sulpiride (10 instead of 20 μM) were co-applied to counteract an increased dopamine level induced by GBR12909. LTD blockade by GBR12909 (Fig. 3A) was cancelled (filled triangle). Representative mean responses taken before the application of GBR12909 + SCH23390 + sulpiride and 45 min after 3-Hz stimuli are presented, showing the cancellation of the LTD blockade. (B) The same cancellation of GBR12909-induced LTD block was achieved by SCH23390 alone (filled diamond). (C) sulpiride alone failed to cancel the GBR12909-induced LTD block (open triangle) (see text for 20 μM sulpiride effect).

Excess Amount of Dopamine Acts on D1 Receptors to Impair LTD

Which subtype of dopamine receptors is involved in the above GBR12909-induced impairment of early LTD? We showed in Figure 2 that, for the normal LTD induction, dopamine must act on both D1-like and D2-like receptors where the role for D2-like receptors, which have higher affinities to dopamine than D1-like (particularly D1) receptors (Missale et al. 1998), appears larger than D1-like receptors. But, under excess levels of dopamine, the balance between D1- and D2-like receptor activation may shift toward D1-like receptors, because the density of D1-like receptors is higher than D2-like receptors in the cortex (Lidow et al. 1991; Hall et al. 1994), despite their low affinities to dopamine (Missale et al. 1998). Hyperstimulation of D1-like receptors as may occur under hyperdopaminergia was indeed shown to be harmful for executive function (Zahrt et al. 1997; Williams and Castner 2006; Vijayraghavan et al. 2007).

To test which subtype of dopamine receptors is involved in the GBR12909-induced LTD impairment, we applied either SCH23390 (1 μM, n = 7) or sulfiride (10 μM, n = 7) in the presence of GBR12909 (5 nM). First, SCH23390 alone still cancelled the GBR12909-induced LTD impairment (−20.9 ± 3.8% 45 min after 3-Hz stimuli, Fig. 4B, filled diamond). There was a significant main effect during the post 3-Hz stimuli period between this SCH23390 + GBR12909 group and the group where LTD was impaired by GBR12909 (1–5 nM; F1,12 = 11.902, P < 0.005). The EPSP decrease 45 min after 3-Hz stimuli in this SCH23390 + GBR12909 group (−20.9 ± 3.8%) was significantly larger than that of the LTD impairment group (−1.7 ± 4.1%; P < 0.005). Secondly, in contrast, sulpiride alone failed to cancel the GBR12909-induced LTD impairment, so that no significant LTD appeared after 3-Hz stimuli in the presence of sulpiride + GBR12909 (−7.4 ± 3.4% at 45 min, Fig. 4C, filled circle). There was no main effect between this group and the GBR12909-induced LTD impairment group during the post 3-Hz stimuli period (F1,12 = 0.026, P > 0.08). The EPSP change 45 min after 3-Hz stimuli in the sulpiride + GBR12909 group was not different from the GBR12909-induced LTD impairment group (−1.7 ± 4.1%; P > 0.2, 2-tailed t-test), but significantly smaller than the normal LTD depicted in Figure 1B (−23.7 ± 6.4%, P < 0.05). To exclude the possibility that GBR12909 still caused excess stimulation of D2-like receptors in the presence of 10 μM sulpiride, we used 20 μM sulpiride in the presence of GBR12909 (5 nM). LTD was still absent in this 20 μM group. The EPSP change 45 min after 3-Hz stimuli showed only −6.0 ± 5.0%, which was significantly smaller than the normal LTD (F1,12 = 5.659, P < 0.04; P < 0.05 by 2-tailed t-test at 45 min). These results indicate that the excess level of extracellular dopamine in the presence of GBR12909 acts mainly on D1-like receptors to impair early LTD.

Impairment of LTD by DAT Inhibition is Cancelled by the Allosteric Modulator of mGluR5

A potentially important issue for clinical neuropsychiatry is to reverse the GBR12909-induced LTD impairment by drugs. In general, stimulation of mGluR5 is an effective way to antagonize behavioral changes induced by psychoactive drugs (Tzavara et al. 2009), and recent data showed that the allosteric-positive modulator of mGluR5 3-cyano-N-(1,3-diphenyl-1H-pyrazol-5-yl)benzamide (CDPPB) inhibited amphetamine-induced hyperlocomotion and deficit of prepulse inhibition (Kinney et al. 2005), as well as MK801-induced increases of PFC neuronal firing and the associated executive cognitive function deficits (Lecourtier et al. 2007; Darrah et al. 2008). Since amphetamine and MK801 increase dopamine concentrations in the PFC (Lecourtier et al. 2007, Pum et al. 2007), it appears worth testing whether the present GBR12909-induced impairment of LTD can be reversed by CDPPB. We thus applied CDPPB [1 μM; chosen based on Lindsley et al. (2004) and Kinney et al. (2005)] 10 min before GBR12909 application (5 nM) until the end of the application/3-Hz stimuli (n = 6). We verified in separate slices that baseline responses do not change in the presence of CDPPB for the duration equivalent to the preconditioning drug perfusion + conditioning stimuli delivery + plasticity monitoring (−3.6 ± 5.2% at the time equivalent to 45 min after 3-Hz stimulation, n = 5, P > 0.1, paired t-test). Under this condition, CDPPB cancelled the GBR12909-induced impairment of LTD so that a normal degree of LTD appeared in the presence of GBR12909 (Fig. 5A, filled diamond). There was a significant main effect during the post 3-Hz stimuli period between the CDPPB + GBR12909 group and GBR12909-induced LTD impairment group (1–5 nM; F1,11 = 7.842, P < 0.02). The EPSP slope change 45 min after 3-Hz stimuli in the CDPPB + GBR12909 group (−23.9 ± 7.5%) was significantly larger than that of the LTD impairment group (−1.7 ± 4.1%; P < 0.02, 2-tailed t-test). Intriguingly, when CDPPB was applied alone, it significantly reduced LTD (−6.2 ± 4.7% at 45 min, n = 6, P < 0.05 compared with the normal LTD group; Fig. 5B, filled triangle). Thus, CDPPB specifically restored early LTD under hyperdopaminergia.

Figure 5.

CDPPB, the positive allosteric modulator of mGluR5, cancelled the GBR12909-induced LTD block. (A) CDPPB (1 μM) coapplied with GBR12909-restored LTD (filled diamond), which was blocked by GBR12909 alone (filled square). The cancellation of LTD blockade is also shown by the mean responses taken before CDPPB + GBR12909 and 45 min after 3-Hz stimuli. (B) Intriguingly, CDPPB applied alone blocked LTD (filled triangle). This group was compared with 3-Hz stimuli-induced LTD (filled square). Representative mean responses taken before CDPPB application and 45 min after 3-Hz stimuli also show the LTD blockade.

Figure 5.

CDPPB, the positive allosteric modulator of mGluR5, cancelled the GBR12909-induced LTD block. (A) CDPPB (1 μM) coapplied with GBR12909-restored LTD (filled diamond), which was blocked by GBR12909 alone (filled square). The cancellation of LTD blockade is also shown by the mean responses taken before CDPPB + GBR12909 and 45 min after 3-Hz stimuli. (B) Intriguingly, CDPPB applied alone blocked LTD (filled triangle). This group was compared with 3-Hz stimuli-induced LTD (filled square). Representative mean responses taken before CDPPB application and 45 min after 3-Hz stimuli also show the LTD blockade.

Molecular Mechanisms of LTD

We next revealed some of the molecular mechanisms underlying the present early LTD. First, the involvement of phospholipase C (PLC) and protein kinase C (PKC) was tested, since a concurrent stimulation of D1- and D2-like receptors as occurs in the present LTD induction (Fig. 2) synergistically recruits PLC and PKC pathways in rat frontal neurons (Lee et al. 2004). The specific PLC inhibitor U73122 (4 μM, n = 6, Fig. 6A, filled diamond; Thompson et al. 1991) or PKC inhibitor RO31-8220 (0.4 μM, n = 6, Fig. 6B, filled circle; Davis et al. 1989) was bath-applied 15 min before 3-Hz stimuli until the end of the stimuli. In separate slices, these drugs alone did not alter baseline responses for the duration equivalent to the preconditioning drug perfusion + conditioning stimuli delivery + plasticity monitoring (changes at the time equivalent to 45 min after 3-Hz stimuli; U73122 group, −1.1 ± 4.5%, n = 5; RO31-8220 group, 0.5 ± 5.1, n = 6; P > 0.1 by paired t-test). Under these conditions, both inhibitors blocked LTD. There was a significant main effect compared with the normal LTD (Fig. 1B) for both of these groups during the post 3-Hz stimuli period (F1,11 = 13.288, P < 0.005 and F1,11 = 5.276, P < 0.05, respectively). The EPSP changes 45 min after 3-Hz stimuli (5.7 ± 5.6 and –0.5 ± 5.9%, respectively) were significantly smaller than the normal LTD (P < 0.005 and <0.02, respectively). Thus, the present early LTD requires the activation of PLC and PKC.

Figure 6.

LTD requires the activation of PLC, PKC, and ERK1/2. (A) The specific PLC inhibitor U73122 (4 μM) blocked 3-Hz stimuli-induced LTD (filled diamond). (B) The specific PKC inhibitor RO31-8220 (0.4 μM) blocked LTD (filled circle). (C) The specific MEK inhibitor PD98059 (20 μM) blocked LTD (filled triangle). (D) Western blot measurement of ERK2 phosphorylation in various pharmacological conditions. There was a significant increase of p-ERK level in 3 Hz + GBR12909 condition (LTD block), which was not seen when D1- and D2-receptor antagonists or CDPPB were copresent with GBR12909 (LTD restoration conditions; see text for statistics). The number of observations was 7 for each group. An example gel showing p-ERK levels in these conditions is also presented. (E) Counteracting the GBR12909-induced p-ERK2 increase (see D) by 5 μM PD98059 restored LTD. LTD blockade by GBR12909 (open circle) was cancelled by this 5 μM PD98059 coapplication (open triangle) so that LTD indistinguishable from the normal LTD occurred under this condition. This cancellation is also shown by the representative mean responses taken before GBR12909 + PD98059 application and 45 min after 3-Hz stimuli.

Figure 6.

LTD requires the activation of PLC, PKC, and ERK1/2. (A) The specific PLC inhibitor U73122 (4 μM) blocked 3-Hz stimuli-induced LTD (filled diamond). (B) The specific PKC inhibitor RO31-8220 (0.4 μM) blocked LTD (filled circle). (C) The specific MEK inhibitor PD98059 (20 μM) blocked LTD (filled triangle). (D) Western blot measurement of ERK2 phosphorylation in various pharmacological conditions. There was a significant increase of p-ERK level in 3 Hz + GBR12909 condition (LTD block), which was not seen when D1- and D2-receptor antagonists or CDPPB were copresent with GBR12909 (LTD restoration conditions; see text for statistics). The number of observations was 7 for each group. An example gel showing p-ERK levels in these conditions is also presented. (E) Counteracting the GBR12909-induced p-ERK2 increase (see D) by 5 μM PD98059 restored LTD. LTD blockade by GBR12909 (open circle) was cancelled by this 5 μM PD98059 coapplication (open triangle) so that LTD indistinguishable from the normal LTD occurred under this condition. This cancellation is also shown by the representative mean responses taken before GBR12909 + PD98059 application and 45 min after 3-Hz stimuli.

We previously showed in the PFC that the induction of dopamine-dependent LTP and LTD by high-frequency stimuli requires increased phosphorylation of ERK1/2 (Otani et al. 1999; Kolomiets et al. 2009). ERK1/2 is involved in long-term cognitive function of the PFC (Runyan et al. 2004) and serves as a gating molecule for plasticity induction in the PFC (Kolomiets et al. 2009; Marzo et al. 2010). We therefore tested the involvement of ERK1/2 in the present early LTD. We applied PD98059 (20 μM; Alessi et al. 1995), the specific inhibitor of mitogen-activated protein kinase–ERK kinase (MEK) that specifically inhibits ERK1/2, 15 min before 3-Hz stimuli until the end of the stimuli (n = 6, Fig. 6C, filled triangle). In separate slices, the baseline responses were not changed by PD98059 for the duration equivalent to the preconditioning drug perfusion + conditioning stimuli delivery + plasticity monitoring (−2.0 ± 5.7% at the time equivalent to 45 min after 3 Hz stimuli, n = 6, P > 0.1 by paired t-test). Under this condition, PD98059 significantly reduced LTD. While there was no main effect between this group and the normal LTD group (depicted in Fig. 1B) during the post 3-Hz stimuli period (F1,11 = 2.809, P < 0.15), there was a highly significant group × time interaction (F99,1089 = 1.830, P < 0.0001). The EPSP slope change 45 min after 3-Hz stimuli in the PD98059 group (−4.8 ± 5.5%, n = 6) was significantly smaller than the normal LTD (−23.7 ± 6.4%; P < 0.05). Thus, ERK1/2 inhibition significantly promoted the decay of LTD, consistent with the fact that ERK1/2 serves for long-term memory function of the PFC (Runyan et al. 2004).

We then quantified the phosphorylated form of ERK1/2 (p-ERK) in the PFC tissue by the use of western blot analysis in order to correlate the degree of ERK1/2 phosphorylation to the above-described early LTD induction, impairment, and restoration. To prepare the prelimbic tissues, experiments were performed identically as in the above electrophysiological experiments except the response monitoring. Stimulus intensity was set at the mean intensity calculated from the last 20 electrophysiological experiments (∼30 μA). Immediately after the end of 3-Hz stimuli, the prelimbic area was dissected on ice and stored in −80°C for later analysis. First, LTD-inducing 3-Hz stimuli increased the p-ERK level (Fig. 6D) albeit to a nonsignificant degree: this was probably because only a limited subpopulation of synapses was affected by the focal synaptic stimuli unlike the bath application of agonists as previously adopted (Kolomiets et al. 2009). But, when GBR12909 (5 nM) was present during the 3-Hz stimuli, p-ERK level increased to a significant level (P < 0.05 compared with the control). This increased p-ERK level was reduced back to a level indistinguishable from the non-GBR12909 + 3-Hz stimuli group when CDPPB (1 μM) was coapplied with GBR12909, that is, the condition where GBR12909-induced LTD impairment was cancelled (Fig. 5A; P > 0.1 compared with the 3-Hz stimuli group, but P < 0.05 compared with the GBR12909 group). Furthermore, when 1 μM SCH23390 and 10 μM sulpiride were coapplied with GBR12909, that is, another condition where GBR12909-induced LTD impairment was cancelled (Fig. 4A), the increased p-ERK1/2 level was again reduced to a level indistinguishable from the 3-Hz stimuli group (P > 0.1). To assure the causal relation between the GBR12909-induced p-ERK increase and the LTD impairment, we conducted further electrophysiological experiments where PD98059 was coapplied at a low concentration to counteract the GBR12909-induced p-ERK increase. Provided with the IC50 value of PD98059 on MEK2 (50 μM; Alessi et al. 1995), we chose the 10 times less concentration (5 μM) instead of 20 μM, which had blocked LTD (Fig. 6C). We found that PD98059 at 5 μM cancelled the GBR12909-induced impairment of LTD and successfully restored LTD (−28.4 ± 5.2%, n = 6, P < 0.005; Fig. 6E, filled triangle), indicating that the increased p-ERK level indeed hampers LTD.

Discussion

Previously, repeated delivery of a brief train of burst stimuli at 1 Hz was shown to induce LTD in hippocampus–PFC synapses in vivo (Takita et al. 1999). In slices, single 3-Hz stimuli induced LTD in the mouse PFC (Huang et al. 2004). The present LTD is thus the first LTD demonstrated in the rat PFC by low-frequency single stimuli. We showed further that this LTD was inducible under intact synaptic inhibition in both juvenile and adult rats and that it survived the presence of tonic background dopamine unlike the form of LTD induced by high-frequency stimuli in the PFC (Matsuda et al. 2006; Kolomiets et al. 2009). This latter form of LTD usually requires a high level of dopamine receptor stimulation by exogenous dopamine and is NMDA receptor-independent (Otani et al. 1998, 1999), similar to the LTD shown in the mouse PFC (Huang et al. 2004). These facts prompted us to indicate that this latter LTD is a dysfunctional form of synaptic depression (Goto et al. 2010). In contrast, the present LTD induced by single stimuli may represent a physiologically relevant synaptic depression in the PFC. We note, however, that we did not assess LTD maintenance longer than 50 min (see Materials and Methods), precluding conclusions on long-lasting forms of LTD (late LTD) that last for several hours (cf. Manahan-Vaughan and Kulla 2003).

Our main finding was that this physiological early LTD was impaired by selective inhibition of the DAT. This finding is consistent with our previous data suggesting the existence of an “inverted-U” shape relation between the extracellular dopamine level and plasticity induction in the rat PFC (Kolomiets et al. 2009; Goto et al. 2010). But in the present study, we moreover showed that the excess level of dopamine by DAT inhibition acts critically on D1-like dopamine receptors to impair LTD, where the precise involvement of D1 versus D5 receptors still remains to be seen. Nevertheless, this novel finding on the D1-like receptor involvement indicates that, for this early LTD induction, the degree of D1-like receptor stimulation determines the inverted-U shape profile, reminiscent of the earlier results that over-stimulation of D1-like receptors in the PFC disrupted executive cognitive function (Zahrt et al. 1997; Williams and Castner 2006; Vijayraghavan et al. 2007) and helped reinstate the CPP (Sanchez et al. 2003).

The above over-stimulation of D1-like receptors under excess levels of dopamine moreover agrees with the recent modeling study that showed that the increased dopamine level upon the burst firing of dopamine neurons leads primarily to an increased occupancy of D1 receptors (Dreyer et al. 2010). Equally, previous experimental observations that a facilitated LTP in DAT-KO mice was caused by hyperstimulation of D1-like receptors are consistent with our present data (Morice et al. 2007; Bai et al. 2010; but also see Swant and Wagner 2007; Xu et al. 2009 for some discrepancies). Furthermore, the critical inverted-U shape type involvement of D1-like receptors has been reported in the rat chronic stress model, where agonist stimulation of D1-like receptors rescued stress-induced working memory deficit caused by a reduced dopamine level in the PFC (Mizoguchi et al. 2000). In addition, the present D1-like receptor-mediated LTD inhibition is reminiscent of the observations made in the hippocampus and visual cortex that 1) D1 receptor stimulation converted spike-timing-dependent LTD to LTP (Zhang et al. 2009), 2) stimulation of Gs-coupled receptors inhibited low-frequency stimuli-induced LTD (Huang et al. 2012), and 3) D1-like receptor agonist reversed low-frequency stimuli-induced LTD (Mockett et al. 2007; but see Lemon and Manahan-Vaughan 2006 for a contradictory in vivo result).

In human motor cortex, on the other hand, a similar nonlinear modulation by dopamine has been shown to exist for neuroplasticity induction. In this case, l-dihydroxyphenylalanine administration exerted an inverted-U shape dose-dependent enhancement of motor cortex neuronal excitability induced by associative stimuli (Thirugnanasambandam et al. 2011). Intriguingly here, no such dose-dependency was detected when a less physiological, continuous DC stimulation method was employed (Monte-Silva et al. 2010). Thus, together with the present animal data, the inverted-U shape modulation of neuroplasticity indeed appears to serve as a mechanistic basis for dopaminergic modulation of the brain function including cognition.

It is important to add here, furthermore, that an elevated DA level affects other neuromodulators such as noradrenaline. Thus, the augmentation of the DA level in the PFC increases the extracellular level of noradrenaline in the PFC through D1 receptor stimulation (Pan et al. 2004). Noradrenaline is another powerful modulator for neuroplasticity induction in the PFC and the hippocampus (Munro et al. 2001; Marzo et al. 2010), and it increases p-ERK in the PFC (Marzo et al. 2010). There may be an inverted-U shape regulation of plasticity induction also by noradrenaline.

As to the molecular mechanism, the present low-frequency stimuli-induced early LTD required the activation of ERK1/2, and an over-increase of p-ERK level by DAT inhibition impaired the LTD (Fig. 6D, E). These results indicate that an inverted-U shape profile may apply also to the relation between ERK phosphorylation and LTD induction. In this regard, we point out that the allosteric mGluR5 potentiator CDPPB appears to exert its de-blocking effect on GBR12909-induced LTD impairment by down-regulating the over-activated ERK by GBR12909 application. It is unlikely indeed that CDPPB rescues LTD through direct intervention on dopaminergic transmission in the PFC (Lecourtier et al. 2007). Currently, more detailed molecular dissection of the CDPPB effects on p-ERK and LTD is unfortunately unavailable. But, we nevertheless point out that a concurrent dopamine receptor stimulation appears necessary for the rescuing effect of CDPPB on LTD, since sole stimulation of mGluR1/5 rather increases the p-ERK level in the PFC (Otani et al. 1999) and since CDPPB alone rather blocked LTD (Fig. 5B).

Our observation that the DAT inhibitor impaired LTD would helps us understand at least part of the cocaine action in the PFC. Acute administration of cocaine disrupts reversal learning (Jentsch et al. 2002) that depends on the intact PFC (De Bruin et al. 2000) and facilitates reinstatement of the CPP mainly through the inhibition of DAT in the PFC (Sanchez et al. 2003; Schmidt and Pierce 2006). These behavioral data suggest that cocaine impairs behavioral flexibility and retains a previously learned rigid goal direction. It then appears plausible that these modifications involve disrupted LTD in the PFC since LTD permits behavioral adaptations when response-reward contingency has changed (Morice et al. 2007; Nicholls et al. 2008). In this regard, the canceling effect of CDPPB on GBR12909-induced LTD impairment is particularly relevant, because CDPPB also counteracts the abnormal neural activity in the PFC seen under hyperdopaminergia and the associated deficit of executive cognitive function (Lecourtier et al. 2007; Darrah et al. 2008).

The above restoration of LTD by CDPPB under hyperdopaminergia is reminiscent of our early observation that mGluR1/5 agonist S-3,5-dihydroxyphenylglycine induced LTD only when it was coapplied with dopamine (Otani et al. 1999). But an intriguing finding here was that when applied alone, CDPPB rather inhibited LTD (Fig. 5B). This was unexpected, since if anything, agonist stimulation of mGluR1/5 leads to LTD (Palmer et al. 1997). We speculate that the difference may derive from the fact that the use-dependent allosteric potentiation of mGluR5 is different from simple agonist stimulation (Kinney et al. 2005). Detailed underlying mechanisms remain to be investigated.

Studies now indicate that LTP- and LTD-like changes indeed underlie memory storage in the brain (Malenka and Bear 2004; Pastalkova et al. 2006; Nicholls et al. 2008). We more recently suggested that aberrant forms of synaptic plasticity exist in the PFC that underlie cognitive impairments seen in psychiatric disorders (Goto et al. 2010). A recent demonstration indeed has strengthened the causality between drug (cocaine)-induced D1 receptor-dependent abnormal plasticity and the behavioral deficits (Pascoli et al. 2012). In the present report, we showed another example of the aberrant modulation of PFC synaptic plasticity, which may help us better understand the cellular basis of rigid goal direction. The involvement of D1 receptor over-stimulation in this abnormality and its reversal by CDPPB may provide us with useful information for the future development of drugs to treat cognitive deficits seen under brain hyperdopaminergia.

Supplementary Material

Supplementary material can be found at: http://www.cercor.oxfordjournals.org/.

Funding

This work was supported by the INSERM, CNRS, UPMC, Nagai Foundation, China Scholarship Council, French Minister of Foreign Affaires, French Minister of Research, and Sophia University ORC grant.

Notes

Conflict of Interest: None declared.

References

Alessi
DR
Cuenda
A
Cohen
P
Dudley
DT
Saltiel
AR
PD 098059 is a specific inhibitor of the activation of mitogen-activated protein kinase kinase in vitro and in vivo
J Biol Chem
 , 
1995
, vol. 
270
 (pg. 
27489
-
27494
)
Andersen
PH
The dopamine uptake inhibitor GBR 12909: selectivity and molecular mechanism of action
Eur J Pharmacol
 , 
1989
, vol. 
166
 (pg. 
493
-
504
)
Artola
A
Singer
W
Long-term potentiation and NMDA receptors in rat visual cortex
Nature
 , 
1987
, vol. 
330
 (pg. 
649
-
652
)
Bai
J
Gorgievski
V
Tzavara
E
Nosten-Bertrand
M
Giros
B
Otani
S
Dopamine transporter (DAT) activity regulates the induction of synaptic plasticity in rodent prefrontal cortex
FENS abstr vol
 , 
2010
, vol. 
5
 pg. 
103
 
Barch
DM
Dowd
EC
Goal representations and motivational drive in schizophrenia: the role of prefrontal-striatal interactions
Schizophr Bull
 , 
2010
, vol. 
36
 (pg. 
919
-
934
)
Davis
PD
Hill
CH
Keech
E
Lawton
G
Nixon
JS
Sedgwick
AD
Wadsworth
J
Westmacott
D
Wilkinson
SE
Potent selective inhibitors of protein kinase C
FEBS Lett
 , 
1989
, vol. 
259
 (pg. 
61
-
63
)
Darrah
JM
Stefani
MR
Moghaddam
B
Interaction of N-methyl-d-aspartate and group 5 metabotropic glutamate receptors on behavioral flexibility using a novel operant set-shift paradigm
Behav Pharmacol
 , 
2008
, vol. 
19
 (pg. 
225
-
234
)
De Bruin
JP
Feenstra
MG
Broersen
LM
Van Leeuwen
M
Arens
C
De Vries
S
Joosten
RN
Role of the prefrontal cortex of the rat in learning and decision making: effects of transient inactivation
Prog Brain Res
 , 
2000
, vol. 
126
 (pg. 
103
-
113
)
DeVito
LM
Eichenbaum
H
Memory for the order of events in specific sequences: contributions of the hippocampus and medial prefrontal cortex
J Neurosci
 , 
2011
, vol. 
31
 (pg. 
3169
-
3175
)
Dreyer
J
Herrik
KF
Berg
RW
Hounsgaard
JD
Influence of phasic and tonic dopamine release on receptor activation
J Neurosci
 , 
2010
, vol. 
30
 (pg. 
14273
-
14283
)
Frantseva
MV
Fitzgerald
PB
Chen
R
Möller
B
Daigle
M
Daskalakis
ZJ
Evidence for impaired long-term potentiation in schizophrenia and its relationship to motor skill learning
Cereb Cortex
 , 
2008
, vol. 
18
 (pg. 
990
-
996
)
Fuster
JM
Memory in the Cerebral Cortex: An Empirical Approach to Neural Networks in the Human and Nonhuman Primate
 , 
1995
Cambridge, MA
The MIT Press
Gainetdinov
RR
Caron
MG
Genetics of childhood disorders: XXIV. ADHD, part 8: hyperdopaminergic mice as an animal model of ADHD
J Am Acad Child Adolesc Psychiatry
 , 
2001
, vol. 
40
 (pg. 
380
-
382
)
Giros
B
Caron
MG
Molecular characterization of the dopamine transporter
Trends Pharmacol Sci
 , 
1993
, vol. 
14
 (pg. 
43
-
49
)
Gorelova
N
Seamans
JK
Yang
CR
Mechanisms of dopamine activation of fast-spiking interneurons that exert inhibition in rat prefrontal cortex
J Neurophysiol
 , 
2002
, vol. 
88
 (pg. 
3150
-
3166
)
Goto
Y
Yang
CR
Otani
S
Functional and dysfunctional synaptic plasticity in prefrontal cortex: roles in psychiatric disorders
Biol Psychiatry
 , 
2010
, vol. 
67
 (pg. 
199
-
207
)
Greenwood
TA
Schork
NJ
Eskin
E
Kelsoe
JR
Identification of additional variants within the human dopamine transporter gene provides further evidence for an association with bipolar disorder in two independent samples
Mol Psychiatry
 , 
2006
, vol. 
11
 (pg. 
125
-
133
)
Gurden
H
Tassin
JP
Jay
TM
Integrity of the mesocortical dopaminergic system is necessary for complete expression of in vivo hippocampal-prefrontal cortex long-term potentiation
Neuroscience
 , 
1999
, vol. 
94
 (pg. 
1019
-
1027
)
Hall
H
Sedvall
G
Magnusson
O
Kopp
J
Halldin
C
Farde
L
Distribution of D1- and D2-dopamine receptors, and dopamine and its metabolites in the human brain
Neuropsychopharmacology
 , 
1994
, vol. 
11
 (pg. 
245
-
256
)
Hirsch
JC
Crepel
F
Use-dependent changes in synaptic efficacy in rat prefrontal neurons in vitro
J Physiol
 , 
1990
, vol. 
427
 (pg. 
31
-
49
)
Huang
S
Treviño
M
He
K
Ardiles
A
de Pasquale
R
Guo
Y
Palacios
A
Huganir
R
Kirkwood
A
Pull-push neuromodulation of LTP and LTD enables bidirectional experience-induced synaptic scaling in visual cortex
Neuron
 , 
2012
, vol. 
73
 (pg. 
497
-
510
)
Huang
YY
Simpson
E
Kellendonk
C
Kandel
ER
Genetic evidence for the bidirectional modulation of synaptic plasticity in the prefrontal cortex by D1 receptors
Proc Natl Acad Sci U S A
 , 
2004
, vol. 
101
 (pg. 
3236
-
3241
)
Jentsch
JD
Olausson
P
De La Garza
R
II
Taylor
JR
Impairments of reversal learning and response perseveration after repeated, intermittent cocaine administrations to monkeys
Neuropsychopharmacology
 , 
2002
, vol. 
26
 (pg. 
183
-
190
)
Kinney
GG
O'Brien
JA
Lemaire
W
Burno
M
Bickel
DJ
Clements
MK
Chen
TB
Wisnoski
DD
Lindsley
CW
Tiller
PR
Jr
, et al.  . 
A novel selective positive allosteric modulator of metabotropic glutamate receptor subtype 5 has in vivo activity and antipsychotic-like effects in rat behavioral models
J Pharmacol Exp Ther
 , 
2005
, vol. 
313
 (pg. 
199
-
206
)
Kolomiets
B
Marzo
A
Caboche
J
Vanhoutte
P
Otani
S
Background dopamine concentration dependently facilitates long-term potentiation in rat prefrontal cortex through postsynaptic activation of extracellular signal-regulated kinases
Cereb Cortex
 , 
2009
, vol. 
19
 (pg. 
2708
-
2718
)
Lecourtier
L
Homayoun
H
Tamagnan
G
Moghaddam
B
Positive allosteric modulation of metabotropic glutamate 5 (mGlu5) receptors reverses N-methyl-d-aspartate antagonist-induced alteration of neuronal firing in prefrontal cortex
Biol Psychiatry
 , 
2007
, vol. 
62
 (pg. 
739
-
746
)
Lee
SP
So
CH
Rashid
AJ
Varghese
G
Cheng
R
Lança
AJ
O'Dowd
BF
George
SR
Dopamine D1 and D2 receptor co-activation generates a novel phospholipase C-mediated calcium signal
J Biol Chem
 , 
2004
, vol. 
279
 (pg. 
35671
-
35678
)
Lemon
N
Manahan-Vaughan
D
Dopamine D1/D5 receptors gate the acquisition of novel information through hippocampal long-term potentiation and long-term depression
J Neurosci
 , 
2006
, vol. 
26
 (pg. 
7723
-
7729
)
Lidow
MS
Goldman-Rakic
PS
Gallager
DW
Rakic
P
Distribution of dopaminergic receptors in the primate cerebral cortex: quantitative autoradiographic analysis using [3H]raclopride, [3H]spiperone and [3H]SCH23390
Neuroscience
 , 
1991
, vol. 
40
 (pg. 
657
-
671
)
Lindsley
CW
Wisnoski
DD
Leister
WH
O'brien
JA
Lemaire
W
Williams
DL
Jr
Burno
M
Sur
C
Kinney
GG
Pettibone
DJ
, et al.  . 
Discovery of positive allosteric modulators for the metabotropic glutamate receptor subtype 5 from a series of N-(1,3-diphenyl-1H- pyrazol-5-yl)benzamides that potentiate receptor function in vivo
J Med Chem
 , 
2004
, vol. 
47
 (pg. 
5825
-
5828
)
Manahan-Vaughan
D
Kulla
A
Regulation of depotentiation and long-term potentiation in the dentate gyrus of freely moving rats by dopamine D2-like receptors
Cereb Cortex
 , 
2003
, vol. 
13
 (pg. 
123
-
135
)
Malenka
RC
Bear
MF
LTP and LTD: an embarrassment of riches
Neuron
 , 
2004
, vol. 
44
 (pg. 
5
-
21
)
Marzo
A
Bai
J
Caboche
J
Vanhoutte
P
Otani
S
Cellular mechanisms of long-term depression induced by noradrenaline in rat prefrontal neurons
Neuroscience
 , 
2010
, vol. 
169
 (pg. 
74
-
86
)
Matsuda
Y
Marzo
A
Otani
S
The presence of background dopamine signal converts long-term synaptic depression to potentiation in rat prefrontal cortex
J Neurosci
 , 
2006
, vol. 
26
 (pg. 
4803
-
4810
)
Missale
C
Nash
SR
Robinson
SW
Jaber
M
Caron
MG
Dopamine receptors: from structure to function
Physiol Rev
 , 
1998
, vol. 
78
 (pg. 
189
-
225
)
Mizoguchi
K
Yuzurihara
M
Ishige
A
Sasaki
H
Chui
DH
Tabira
T
Chronic stress induces impairment of spatial working memory because of prefrontal dopaminergic dysfunction
J Neurosci
 , 
2000
, vol. 
20
 (pg. 
1568
-
15674
)
Mockett
BG
Guevremont
D
Williams
JM
Abraham
WC
Dopamine D1/D5 receptor activation reverses NMDA receptor-dependent long-term depression in rat hippocampus
J Neurosci
 , 
2007
, vol. 
27
 (pg. 
2918
-
2926
)
Monte-Silva
K
Liebetanz
D
Grundey
J
Paulus
W
Nitsche
MA
Dosage-dependent non-linear effect of L-DOPA on human motor cortex plasticity
J Physiol
 , 
2010
, vol. 
588
 (pg. 
3415
-
3424
)
Morice
E
Billard
JM
Denis
C
Mathieu
F
Betancur
C
Epelbaum
J
Giros
B
Nosten-Bertrand
M
Parallel loss of hippocampal LTD and cognitive flexibility in a genetic model of hyperdopaminergia
Neuropsychopharmacology
 , 
2007
, vol. 
32
 (pg. 
2108
-
2116
)
Morris
SH
Knevett
S
Lerner
EG
Bindman
LJ
Group I mGluR agonist DHPG facilitates the induction of LTP in rat prelimbic cortex in vitro
J Neurophysiol
 , 
1999
, vol. 
82
 (pg. 
1927
-
1933
)
Mundorf
ML
Joseph
JD
Austin
CM
Caron
MG
Wightman
RM
Catecholamine release and uptake in the mouse prefrontal cortex
J Neurochem
 , 
2001
, vol. 
79
 (pg. 
130
-
142
)
Munro
CA
Walling
SG
Evans
JH
Harley
CW
Beta-adrenergic blockade in the dentate gyrus in vivo prevents high frequency-induced long-term potentiation of EPSP slope, but not long-term potentiation of population spike amplitude
Hippocampus
 , 
2001
, vol. 
11
 (pg. 
322
-
328
)
Nicholls
RE
Alarcon
JM
Malleret
G
Carroll
RC
Grody
M
Vronskaya
S
Kandel
ER
Transgenic mice lacking NMDAR-dependent LTD exhibit deficits in behavioral flexibility
Neuron
 , 
2008
, vol. 
58
 (pg. 
104
-
117
)
Otani
S
Auclair
N
Desce
JM
Roisin
M-P
Crépel
F
Dopamine receptors and groups I and II mGluRs cooperate for long-term depression induction in rat prefrontal cortex through converging postsynaptic activation of MAP kinases
J Neurosci
 , 
1999
, vol. 
19
 (pg. 
9788
-
9802
)
Otani
S
Blond
O
Desce
J.-M
Crépel
F
Dopamine facilitates long-term depression of glutamatergic transmission in rat prefrontal cortex
Neuroscience
 , 
1998
, vol. 
85
 (pg. 
669
-
676
)
Otani
S
Connor
JA
Long-term depression of naïve synapses in adult hippocampus induced by asynchronous synaptic activity
J Neurophysiol
 , 
1995
, vol. 
73
 (pg. 
2596
-
2601
)
Palmer
MJ
Irving
AJ
Seabrook
GR
Jane
DE
Collingridge
GL
The group I mGlu receptor agonist DHPG induces a novel form of LTD in the CA1 region of the hippocampus
Neuropharmacology
 , 
1997
, vol. 
36
 (pg. 
1517
-
1532
)
Pan
WHT
Yang
S-Y
Lin
S-K
Neurochemical interaction between dopaminergic and noradrenergic neurons in the medial prefrontal cortex
Synapse
 , 
2004
, vol. 
53
 (pg. 
44
-
52
)
Pascoli
V
Turiault
M
Lüscher
C
Reversal of cocaine-evoked synaptic potentiation resets drug-induced adaptive behavior
Nature
 , 
2012
, vol. 
481
 (pg. 
71
-
75
)
Pastalkova
E
Serrano
P
Pinkhasova
D
Wallace
E
Fenton
AA
Sacktor
TC
Storage of spatial information by the maintenance mechanism of LTP
Science
 , 
2006
, vol. 
313
 (pg. 
1141
-
1144
)
Perona
MT
Waters
S
Hall
FS
Sora
I
Lesch
KP
Murphy
DL
Caron
M
Uhl
GR
Animal models of depression in dopamine, serotonin, and norepinephrine transporter knockout mice: prominent effects of dopamine transporter deletions
Behav Pharmacol
 , 
2008
, vol. 
19
 (pg. 
566
-
574
)
Pum
M
Carey
RJ
Huston
JP
Müller
CP
Dissociating effects of cocaine and d-amphetamine on dopamine and serotonin in the perirhinal, entorhinal, and prefrontal cortex of freely moving rats
Psychopharmacology (Berl)
 , 
2007
, vol. 
193
 (pg. 
375
-
390
)
Reith
MEA
Coffey
LL
Xu
C
Chen
N-H
GBR12909 and 12935 block dopamine uptake into brain synaptic vesicles as well as nerve endings
Eur J Pharmacol
 , 
1994
, vol. 
253
 (pg. 
175
-
178
)
Reith
MEA
Xu
C
Chen
NH
Pharmacology and regulation of the neuronal dopamine transporter
Eur J Pharmacol
 , 
1997
, vol. 
324
 (pg. 
1
-
10
)
Runyan
JD
Moore
AN
Dash
PK
A role for prefrontal cortex in memory storage for trace fear conditioning
J Neurosci
 , 
2004
, vol. 
24
 (pg. 
1288
-
1295
)
Sanchez
CJ
Bailie
TM
Wu
W-R
Liand
N
Sorg
BA
Manipulation of dopamine D1-like receptor activation in the rat medial prefrontal cortex alters stress- and cocaine-induced reinstatement of conditioned place preference behavior
Neuroscience
 , 
2003
, vol. 
119
 (pg. 
497
-
505
)
Schmidt
HD
Pierce
RC
Systemic administration of a dopamine, but not a serotonin or norepinephrine, transporter inhibitor reinstates cocaine seeking in the rat
Behav Brain Res
 , 
2006
, vol. 
175
 (pg. 
189
-
194
)
Swant
J
Wagner
JJ
Dopamine transporter blockade increases LTP in the CA1 region of the rat hippocampus via actiation of the D3 dopamine receptor
Learn Memory
 , 
2007
, vol. 
13
 (pg. 
161
-
167
)
Takita
M
Izaki
Y
Jay
TM
Kaneko
H
Suzuki
SS
Induction of stable long-term depression in vivo in the hippocampal-prefrontal cortex pathway
Eur J Neurosci
 , 
1999
, vol. 
11
 (pg. 
4145
-
4148
)
Thirugnanasambandam
N
Grundey
J
Paulus
W
Nitsche
MA
Dose-dependent nonlinear effect of L-DOPA on paired associative stimulation-induced neuroplasticity in humans
J Neurosci
 , 
2011
, vol. 
31
 (pg. 
5294
-
5299
)
Thompson
AK
Mostafapour
SP
Denlinger
LC
Bleasdale
JE
Fisher
SK
The aminosteroid U-73122 inhibits muscarinic receptor sequestration and phosphoinositide hydrolysis in SK-N-SH neuroblastoma cells. A role for Gp in receptor compartmentation
J Biol Chem
 , 
1991
, vol. 
266
 (pg. 
23856
-
23862
)
Tzavara
ET
Degroot
A
Wade
MR
Davis
RJ
Nomikos
GG
CB1 receptor knockout mice are hyporesponsive to the behavior-stimulating actions of d-amphetamine: role of mGlu5 receptors
Eur Neuropsychopharmacol
 , 
2009
, vol. 
19
 (pg. 
196
-
204
)
Vijayraghavan
S
Wang
M
Birnbaum
SG
Williams
GV
Arnsten
AFT
Inverted-U dopamine D1 receptor actions on prefrontal neurons engaged in working memory
Nat Neurosci
 , 
2007
, vol. 
10
 (pg. 
376
-
384
)
Wigström
H
Gustafsson
B
Facilitated induction of hippocampal long-lasting potentiation during blockade of inhibition
Nature
 , 
1983
, vol. 
301
 (pg. 
603
-
604
)
Williams
GV
Castner
SA
Under the curve: critical issues for elucidating D1 receptor function in working memory
Neuroscience
 , 
2006
, vol. 
139
 (pg. 
263
-
276
)
Xu
T-X
Sotnikova
TD
Liang
C
Zhang
J
Jung
JU
Spealman
RD
Gainetdinov
RR
Yao
W-D
Hyperdopaminergic tone erodes prefrontal long-term potentiation via a D2 receptor-operated protein phosphatase gate
J Neurosci
 , 
2009
, vol. 
29
 (pg. 
14086
-
14099
)
Young
CE
Yang
CR
Dopamine D1-like receptor modulates layer- and frequency-specific short-term synaptic plasticity in rat prefrontal cortical neurons
Eur J Neurosci
 , 
2005
, vol. 
21
 (pg. 
3310
-
3320
)
Zahrt
J
Taylor
JR
Mathew
RG
Arnsten
AF
Supranormal stimulation of D1 dopamine receptors in the rodent prefrontal cortex impairs spatial working memory performance
J Neurosci
 , 
1997
, vol. 
17
 (pg. 
8528
-
8535
)
Zhang
J-C
Lau
P-M
Bi
G-Q
Gain in sensitivity and loss in temporal contrast of STDP by dopaminergic modulation at hippocampal synapses
Proc Natl Acad Sci U S A
 , 
2009
, vol. 
106
 (pg. 
13028
-
13033
)