INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The nucleus accumbens (NAcc) is thought to serve as an interface between limbic and motor systems (Mogenson et al. 1980) and is involved in a number of processes including motivation (Mogenson et al. 1980), attention (Solomon and Staton 1982; van den Bos et al. 1991), and reward (Apicella et al. 1991; Colle and Wise 1988). Limbic innervation of the NAcc includes substantial glutamatergic input from the hippocampus, amygdala, and medial prefrontal cortex (Phillipson and Griffiths 1985). These excitatory inputs have been demonstrated to play an important role in the neurophysiology of the NAcc (Goto and O'Donnell 2001; O'Donnell and Grace 1995). Furthermore, it has been hypothesized that dysregulation of glutamatergic input to the NAcc may underlie the development of psychiatric disorders such as schizophrenia (O'Donnell and Grace 1998) and drug addiction (Wolf 1998). However, the receptor mechanisms responsible for the processing of glutamatergic input within the NAcc have not been fully characterized.

Glutamate, the primary excitatory neurotransmitter in the CNS, activates three major subclasses of ionotropic receptors: N-methyl-D-aspartate (NMDA), -amino-3-hydroxy-5-methyl-4-isoxazole propionate (AMPA), and kainate (KA). Studies that have examined the function of glutamate receptors in the NAcc suggest that the excitatory effects of glutamate, in vivo, are primarily mediated by non-NMDA type (AMPA/KA) receptors (Hu and White 1996; Pennartz et al. 1991). Historically, it has been difficult to distinguish KA receptor (KA-R) from AMPA receptor (AMPA-R) function due to a lack of selective agonists and antagonists. The recent development of selective AMPA-R antagonists has resulted in a number of studies that have begun to unravel the functional role of KA-Rs within the mammalian CNS (for review see Ben-Ari and Cossart 2000; Frerking and Nicoll 2000). To date, KA-Rs have been shown to subserve a traditional postsynaptic role, gating synaptic excitation, in only a limited number of brain regions (Castillo et al. 1997; Frerking et al. 1998; Li and Rogawski 1998; Vignes and Collingridge 1997). However, KA-Rs have also been shown to function as presynaptic receptors at many synapses (Chittajallu et al. 1996; Chergui et al. 2000; Clarke et al. 1997; Frerking et al. 2001; Kamiya and Ozawa 1998, 2000; Kerchner et al. 2001; Rodriguez-Moreno et al. 1997; Schmitz et al. 2000, 2001; Vignes et al. 1998). Activation of these presynaptic KA-Rs potently modulates both glutamate and -aminobutyric acid (GABA) release. Thus if functional KA-Rs are present in the NAcc, they may play an important role in regulating excitatory and inhibitory activity within this brain region.

The aim of the current study was to identify functional KA-Rs within the NAcc and to determine the role of KA-Rs in regulating excitatory activity within this nucleus. Using the whole cell patch-clamp technique, we demonstrate that functional KA-Rs are present on neurons within the NAcc core. These receptors can be activated by exogenous application of KA but do not contribute to excitatory postsynaptic currents (EPSCs) elicited by individual stimuli. Our data also suggest that activation of KA-Rs in the NAcc potently inhibits glutamatergic synaptic transmission via a presynaptic mechanism.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Slice preparation

Male Sprague-Dawley rats (20-40 days old) were anesthetized with halothane and killed by decapitation using a protocol approved by the ACUC of Wake Forest University School of Medicine. Coronal NAcc slices (400 µm) were prepared using a vibrating tissue slicer (Leica VT1000S; Vashaw Scientific, Atlanta, GA). Slices were then maintained at room temperature in oxygenated artificial cerebrospinal fluid (ACSF) containing (in mM) 124 NaCl, 3.3 KCl, 2.4 MgCl₂, 2.5 CaCl₂, 1.2 KH₂PO₄, 10 D-glucose, and 25.9 NaHCO_3, saturated with 95% O₂-5% CO₂. During recordings, slices were perfused with oxygenated ACSF at a flow rate of 2 ml/min. All drugs were applied through the ACSF in known concentrations via calibrated syringe pumps (Razel; Stamford, CT).

Patch-clamp recordings

Methods for whole cell recordings were similar to those reported previously (Weiner et al. 1999). Briefly, electrodes were prepared from filamented borosilicate glass capillary tubes (ID: 0.86 mm) using a horizontal micropipette puller (Sutter P-97; Sutter, Novato, CA). Electrodes were filled with a recording solution containing (in mM) 130 K-gluconate, 10 KCl₂, 5 N-(2,6-dimethyl-phenylcarbamoylmethyl)-triethylammonium bromide (QX-314), 1 ethylene glycol-bis-(-aminoethyl ether)-N,N,N',N'-tetraacetic acid (EGTA), 0.1 CaCl₂, 2 Mg-ATP, 0.2 tris-GTP, and 10 HEPES (free acid). Whole cell patch-clamp recordings were made at room temperature from NAcc core neurons voltage clamped at 70 mV (for AMPA EPSCs) or 20 to 40 mV (for NMDA EPSCs). Recording electrodes were placed ventral to the anterior commissure within the core of the NAcc. Unless otherwise indicated, synaptic currents were evoked every 20 s by electrical stimulation (0.2 ms duration) of tissue adjacent to the recording electrode using a concentric bipolar stimulating electrode (FHC, Bowdoinham, ME). In one experiment, AMPA (10 µM) was applied directly to the soma of NAcc neurons using a Picospritzer III (General Valve, Fairfield, NJ). Recordings were acquired with an Axoclamp 2B amplifier, digitized (Digidata 1200B; Axon Instruments, Foster City, CA), and analyzed on- and off-line using an IBM compatible PC computer and pClamp 8.0 software (Axon Instruments, Foster City, CA).

Pharmacological isolation

Drugs used in the pharmacological isolation of evoked currents included the NMDA receptor antagonist DL-()-2-amino-5-phosphonovaleric acid (APV), the GABA_A receptor antagonist bicuculline methbromide (BIC), the AMPA/KA receptor antagonists 1,2,3,4-tetrahydro-6-nitro-2,3-dioxobenzo[f]quinoxaline-7-sulfonamide (NBQX) and 6,7-dinitroquinoxaline-2,3-dione (DNQX); and the nonselective adenosine receptor antagonist theophylline (all from Sigma, St. Louis, MO). DNQX and NBQX were made up as stock solutions in dimethyl sulfoxide (DMSO; final total concentration of DMSO was <0.05%). APV, BIC, and theophylline were made up as stock solutions in deionized water. The metabotropic glutamate receptor (mGluR) type I/II antagonist (RS)--methyl-4-carboxyphenylglycine (MCPG), the mGluR type III antagonist -cyclopropyl-4-phosphonophenylglycine (CPPG), and the GABA_B receptor antagonist (SCH 50911) were also used (all from Tocris, Bristol, UK). MCPG was made up as a stock solution in 0.1 N NaOH and diluted in ACSF. Both CPPG and SCH 50911 were made up as stock solutions in deionized water.

Statistics

KA effects on the amplitude of EPSCs were defined as percent of control (predrug) values. Agonist-induced inward currents were defined as a difference from control holding current values (i.e., drug value  predrug value) expressed in picoamperes. Either one-way ANOVA or Student's t-tests were then used to analyze these data. When appropriate, post hoc analyses were performed using Dunnett's for comparing multiple groups to control, and Newman-Keuls for all pair-wise comparisons, with significance set at P < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Functional kainate receptors are present on postsynaptic neurons in the NAcc

Although KA-R subunits, including GluR 6/7, and to a lesser extent KA2, have been shown to be expressed in both rodent (Bischoff et al. 1997) and primate (Charara et al. 1999; Kieval et al. 2001) ventral striatum, direct evidence of functional KA-Rs in this brain region is lacking. To test the hypothesis that KA-R subunits expressed in the ventral striatum form functional receptors on NAcc neurons, we measured the amplitude of inward currents evoked by bath application of KA (0.1-1 µM) in the presence of 50 µM APV and 20 µM BIC. Bath application of KA resulted in a concentration-dependent inward current in all cells tested (Fig. 1B). The inward currents evoked by the highest concentration of KA (1 µM) were not significantly attenuated by pretreating the slices with 1 µM NBQX (Fig. 1A). This concentration of NBQX has been shown to maximally block AMPA-Rs without having any effect on KA-Rs in mouse hippocampal slices (Bureau et al. 1999). To ensure that NBQX was also selective for AMPA-Rs over KA-Rs in the rat NAcc, we tested the effect of 1 µM NBQX on AMPA-induced inward currents. Bath application of 5 µM AMPA produced inward currents of similar amplitude to those induced by the application of 1 µM KA (Fig. 1A). Pretreatment of ventral striatal slices with 1 µM NBQX completely blocked AMPA-induced inward currents (Fig. 1A). Thus 1 µM NBQX selectively blocked AMPA-R, but not KA-R function in NAcc neurons. Although KA currents were not altered under conditions where AMPA-Rs were selectively blocked, these currents were completely inhibited by a saturating concentration of the nonselective AMPA/KA receptor antagonist, DNQX (80 µM; Fig. 1A).

View larger version (24K): [in this window] [in a new window]   Fig. 1. Functional kainate receptors (KA-Rs) are present on nucleus accumbens (NAcc) core neurons. A: bar graph summarizing the average amplitude (pA ± SE) of inward currents induced by the bath application of 1 µM KA or 5 µM -amino-3-hydroxy-5-methyl-4 isoxazole propionate (AMPA) in NAcc neurons voltage clamped at 70 mV. The agonist-induced currents were evoked in the presence of 50 µM DL-()-2-amino-5-phosphonovaleric acid (APV), 1 µM 1,2,3,4-tetrahydro-6-nitro-2,3-dioxobenzo[f]quinoxaline-7-sulfonamide (NBQX) or 80 µM 6,7-dinitroquinoxaline-2,3-dione (DNQX). The representative time courses presented above the graph show the inward current elicited by 1 µM KA in the presence of 50 µM APV, 1 µM NBQX, or 80 µM DNQX, and by 5 µM AMPA in the presence of 50 µM APV, and 1 µM NBQX. All recordings were done in the presence of 20 µM bicuculline (BIC). B: bar graph shows the average amplitude (in pA ± SE) of inward currents evoked by bath application of KA (0.05-1 µM). C: non-N-methyl-D-aspartate (non-NMDA) excitatory postsynaptic currents (EPSCs) evoked by single pulse stimulation in the NAcc core are mediated solely by AMPA-Rs. Bar graph illustrates the effect of 1 µM NBQX on non-NMDA EPSCs, as percent of control ± SE, in the presence of 50 µM APV and 20 µM BIC. Traces above the graph are averages of 7 non-NMDA EPSCs evoked in the absence and presence of NBQX (labeled 50 µM APV and 1 µM NBQX, respectively). Numbers in parentheses indicate the number of cells tested under each condition; significant difference from control: * P < 0.05 and ** P < 0.001; significant difference from agonist-induced current: # P < 0.001.

We next tested whether or not KA-Rs contributed to non-NMDA EPSCs in the NAcc core. Non-NMDA EPSCs were elicited by single pulse stimulation of glutamatergic afferents near the recorded neuron in the presence of 50 µM APV and 20 µM BIC. This stimulation protocol evoked a fast inward postsynaptic current in all cells. These non-NMDA EPSCs were completely inhibited by selectively blocking AMPA-Rs with 1 µM NBQX (Fig. 1C).

Kainate inhibits evoked EPSCs

We also sought to determine whether activation of KA-Rs in the NAcc core could alter excitatory synaptic transmission, as recently shown in the hippocampus (Chittajallu et al. 1996; Contractor et al. 2000; Frerking et al. 2001; Kamiya and Ozawa 1998, 2000; Schmitz et al. 2001; Vignes et al. 1998). This hypothesis was tested by evaluating the effect of bath application of KA on the amplitude of pharmacologically isolated AMPA-R- or NMDA-R-mediated EPSCs recorded from NAcc core neurons.

AMPA-R-mediated EPSCs were evoked every 20 s with individual stimuli in the presence of 50 µM APV and 20 µM BIC. A 7-min bath application of KA (0.25-1 µM) significantly decreased the amplitude of AMPA EPSCs in all cells tested (Fig. 2A). This inhibition was concentration dependent, with the highest concentration of KA (1 µM) producing an 86.0 ± 4.4% (mean ± SE) inhibition of AMPA EPSCs, and reversed on wash out of KA.

View larger version (28K): [in this window] [in a new window]   Fig. 2. KA-R activation inhibits glutamatergic synaptic transmission in the NAcc core. A1: time course of KA-R-mediated inhibition of AMPA EPSC amplitude (pA). Traces above the graph are averages of 7 EPSCs evoked at the times indicated by the corresponding letters in the graph. A2: summary bar graph showing the effect of bath application of KA (0.1-1 µM) on AMPA EPSC amplitude as percent of control ± SE. B1: time course of KA-R-mediated inhibition of NMDA EPSC amplitude (pA) recorded in the presence of 1 µM NBQX and 20 µM BIC. Traces above the graph are averages of 7 EPSCs evoked at the times indicated by the corresponding letters in the graph. B2: bar graph summarizes the effect of bath application of 1 µM KA on NMDA EPSCs in the absence and presence of 80 µM DNQX as percent of control ± SE. Symbols are identical to those presented in Fig. 1, with the exception that # indicates significant difference from 1 µM KA + 1 µM NBQX, P < 0.001.

Since KA is an agonist at both KA-R and AMPA-Rs, we also characterized the effect of KA on glutamatergic synaptic transmission under conditions where AMPA-Rs were blocked by assessing the effect of KA on NMDA EPSCs. Pharmacologically isolated NMDA EPSCs were recorded in the presence of 20 µM BIC and an AMPA-R-selective concentration of NBQX (Fig. 2B). Similar to the effect of KA on AMPA EPSCs, a 7-min bath application of 1 µM KA resulted in a significant inhibition of NMDA EPSC amplitude (60.0 ± 9.5% inhibition; Fig. 2B). Pretreating slices with the nonselective AMPA/KA receptor antagonist, DNQX, blocked the inhibitory effect of KA on NMDA EPSCs (Fig. 2B).

Kainate inhibits glutamatergic synaptic transmission via a presynaptic mechanism

To determine whether KA inhibited glutamatergic synaptic transmission in the NAcc via a pre- or postsynaptic mechanism, we performed two experiments. In the first experiment we examined the effect of KA on nonsynaptic glutamatergic currents recorded from NAcc neurons. Currents were evoked every 60 s by local pressure application of 10 µM AMPA. These experiments were carried out in the continuous presence of 500 nM TTX to prevent the possible contribution of synaptic activity to the AMPA-evoked currents. A 7-min bath application of KA had a modest inhibitory effect on AMPA-evoked currents, reducing their amplitude by 26.2 ± 6.0% (Fig. 3). However, this inhibition was significantly less than the antagonism of AMPA EPSCs produced by the same concentration of KA recorded under similar experimental conditions.

View larger version (13K): [in this window] [in a new window]   Fig. 3. KA-R activation does not inhibit nonsynaptic AMPA-R currents to the same extent as it inhibits AMPA-R-mediated EPSCs. A: representative traces demonstrating AMPA-R currents elicited by direct pressure application of 10 µM AMPA in the absence and presence of 1 µM KA. Traces are averages of 4 currents evoked by AMPA under baseline and KA conditions. B: summary bar graph comparing the inhibitory effect of 1 µM KA on AMPA-R-mediated EPSCs (data from Fig. 2) and on currents evoked by local pressure application of 10 µM AMPA. Symbols are identical to those presented in Fig. 1 with the exception that # indicates significant difference from AMPA EPSC, P < 0.001.

In the second experiment, we examined whether KA-R activation altered the release probability for glutamate by determining the effect of KA on paired-pulse facilitation (PPF) of NMDA EPSCs. Two stimuli were paired with an interstimulus interval of 45 ms such that the second EPSC (Peak 2) was potentiated by the first EPSC (Peak 1). We then calculated the ratio of Peak 2/Peak 1 in the absence and presence of 0.5 to 1 µM KA. In all cells tested, both Peak 1 and Peak 2 were inhibited by the application of KA; however, Peak 1 was consistently inhibited to a greater extent. Thus bath application of KA significantly increased PPF (Fig. 4). Notably in three of eight cells tested, KA had a biphasic effect on Peak 2, potentiating Peak 2 amplitude during the first few minutes and then inhibiting it during the latter phase of the KA application. This biphasic effect of KA was never observed with Peak 1 in these paired-pulse experiments or with single EPSCs in any of the other experiments characterizing the presynaptic effects of KA.

View larger version (20K): [in this window] [in a new window]   Fig. 4. KA-R activation increases paired-pulse facilitation (PPF) at glutamatergic synapses in the NAcc core. A: time course showing the increase in PPF of NMDA EPSCs in response to bath application of 0.5 µM KA. Averaged traces above the graph represent paired NMDA EPSCs evoked at the times indicated by the corresponding letters in the graph. B: bar graph summarizing the PPF of NMDA EPSCs, expressed as the mean ratio of P2/P1 ± SE, in the absence and presence of KA. Symbols are identical to those presented in Fig. 1.

Kainate inhibition of evoked EPSCs does not require indirect activation of mGluRs, GABA_B, or adenosine receptors

While the data presented so far are consistent with a presynaptic mechanism for KA inhibition of EPSCs in the NAcc, it does not necessarily follow that KA-Rs are localized to the presynaptic terminals of these synapses. A number of other receptor systems have been hypothesized to indirectly mediate the presynaptic actions of KA, including mGluRs, GABA_B receptors, and adenosine receptors (Chergui et al. 2000; Frerking et al. 1999; Rodriguez-Moreno and Lerma 1998; Schmitz et al. 2001). For this reason, we sought to determine whether the KA-mediated inhibition of glutamatergic synaptic transmission, observed in the NAcc, was due to the secondary activation of these other receptor systems. To test this hypothesis, we pretreated ventral striatal slices with an mGluR antagonist cocktail (1 mM MCPG and 100 µM CPPG), a GABA_B receptor antagonist (20 µM SCH 50911), or a nonselective adenosine receptor antagonist (200 µM theophylline) for 10 min prior to the bath application of KA. Pretreatment with MCPG/CPPG, SCH 50911, or theophylline had no significant effect on KA inhibition of glutamatergic synaptic transmission in the NAcc (Fig. 5).

View larger version (20K): [in this window] [in a new window]   Fig. 5. KA-R-mediated inhibition of synaptic transmission at glutamatergic synapses in the NAcc core is not due to the secondary activation of presynaptic mGlu, GABAB, or adenosine receptors. Bar graph summarizing the effect of 1 µm KA on AMPA EPSC amplitude, expressed as percent of control ± SE, in the absence and presence of the mGluR antagonists (RS)--methyl-4-carboxyphenylglycine (MCPG) and -cyclopropyl-4-phosphonophenylglycine (CPPG; 1 mM and 100 µM, respectively), the GABAB antagonist SCH 50911 (20 µM), or the adenosine receptor antagonist theophylline (200 µM). Superimposed traces presented above the graph are averages of 7 AMPA EPSCs showing the effect of 1 µM KA in the absence and presence of the antagonists. All traces were evoked in the presence of 50 µM APV and 20 µM BIC. Symbols are identical to those presented in Fig. 1.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This study sought to determine whether or not functional KA-Rs are present within the NAcc and to carry out an initial assessment of the physiological role of these receptors in regulating excitatory activity in this brain region. Our results provide the first demonstration that functional KA-Rs are expressed on neurons within the core of the NAcc. We found that these receptors can be activated by exogenous application of KA but they do not contribute to non-NMDA EPSCs evoked by individual stimuli. In addition, we present preliminary evidence suggesting that KA-Rs exert a potent presynaptic inhibitory effect on glutamatergic synaptic transmission in the NAcc.

In our first experiment, bath application of KA evoked an inward current in all NAcc core neurons recorded. Although KA is an agonist at both AMPA-Rs and KA-Rs, three lines of evidence suggest that these inward currents were mediated solely by the activation of KA-Rs. First, the concentrations of KA used in this study have previously been shown to selectively activate KA-Rs (Cossart et al. 1998). Second, a concentration of NBQX, that completely blocked AMPA-induced currents in the NAcc, and has been shown to be selective for AMPA-Rs in other native tissue preparations (Bureau et al. 1999), had no significant effect on KA currents in the NAcc. Finally, these KA-evoked currents were completely blocked by the mixed AMPA/KA receptor antagonist, DNQX. In the absence of commercially available selective KA-R antagonists, similar protocols have been used to demonstrate KA-R function in other brain regions (Frerking et al. 2001).

Despite the presence of agonist-induced KA currents in all NAcc cells tested, KA-Rs were not found to contribute to glutamatergic EPSCs evoked by individual stimuli in this study. This is consistent with studies in the CA1 region of the hippocampus and in the dorsal striatum, which demonstrated functional KA-Rs on postsynaptic cells but did not detect KA-R-mediated EPSCs (Bureau et al. 1999; Chergui et al. 2000). Although our data suggest that KA-Rs may be located extrasynaptically on NAcc neurons, the inability to elicit KA-R EPSCs in response to single stimuli does not rule out the possibility of synaptic KA-Rs in the NAcc. Evoking KA-R EPSCs in other brain regions has been shown to require the use of stimulus trains (Castillo et al. 1997; Li and Rogawski 1998; Vignes and Collingridge 1997). It will be important in future studies to determine whether higher frequency stimulation protocols can evoke KA-R mediated EPSCs in the NAcc.

In concordance with several studies conducted in the hippocampus (Contractor et al. 2000; Kamiya and Ozawa 1998, 2000; Schmitz et al. 2000), our data demonstrate that bath application of KA inhibits both AMPA-R- and NMDA-R-mediated EPSCs in the NAcc core. We also show that this inhibition requires KA-R activation by demonstrating that the effect was insensitive to AMPA-R inhibition but was completely blocked by mixed AMPA/KA-R antagonism. Overall, the effect of KA-R activation on AMPA and NMDA EPSCs was very similar. However, KA did have a slightly greater inhibitory effect on AMPA EPSCs. This difference may have been due to a relatively greater influence of postsynaptic shunting on AMPA EPSCs. Alternatively, the low concentration of NBQX used to isolate NMDA EPSCs may have partially inhibited the KA-Rs responsible for the inhibition of these responses.

We next sought to evaluate the mechanism(s) underlying KA-R-mediated inhibition of glutamatergic synaptic transmission in the NAcc. There is mounting evidence that the inhibition of synaptic transmission by KA in other brain regions is mediated, at least in part, by the activation of presynaptic KA-Rs (Contractor et al. 2000; Frerking et al. 2001; Kamiya and Ozawa 1998, 2000; Schmitz et al. 2000, 2001). While KA inhibition of glutamatergic EPSCs in this study showed a similar concentration dependence to the presynaptic inhibition characterized in other studies (see Schmitz et al. 2001), a postsynaptic mechanism could also account for this effect. For example, it has been argued that KA-induced inward currents in the postsynaptic neuron could shunt I/EPSCs (as discussed in Contractor et al. 2000; Frerking et al. 1999). Although KA does induce inward currents in NAcc neurons, it is unlikely that a postsynaptic mechanism associated with shunting of incoming EPSCs is sufficient to account for the majority of the observed inhibition of glutamatergic EPSCs. Under our recording conditions, 1 µM KA almost completely inhibited AMPA EPSCs evoked in NAcc neurons; however, this same concentration of KA had only a modest inhibitory effect on currents evoked by local pressure application of AMPA. These data suggest that the changes in the passive membrane properties of NAcc neurons induced by KA likely account for only a small component of the inhibition of EPSCs observed in the current study. This result is in agreement with studies in the hippocampus that show that postsynaptic shunting does not account for all of the KA-R-mediated inhibition of synaptic transmission at GABAergic synapses (Frerking et al. 1999). To further support the hypothesis that KA-R-mediated inhibition of glutamatergic synaptic transmission in NAcc neurons was due to a presynaptic action of KA, we examined the effect of KA on PPF. KA inhibition of EPSCs was associated with a significant increase in PPF in these cells. Increases in PPF have been shown to be a reliable indicator of a presynaptic decrease in the probability of neurotransmitter release (Manabe et al. 1993). Although this latter experiment does not exclude the contribution of a postsynaptic component to the inhibitory effect of KA on NAcc EPSCs, it further suggests that this inhibition results, at least in part, from a decrease in the release probability for glutamate.

The inhibitory effect of KA at glutamatergic synapses in the NAcc and other brain regions (Bortolotto et al. 1999; Frerking et al. 2001; Kamiya and Ozawa 1998, 2000; Vignes et al. 1998) may seem somewhat paradoxical in light of the known excitatory effects associated with KA-R activation (see Ben-Ari and Cossart 2000 for review). However, recent studies conducted in the hippocampus suggest that, under some experimental conditions, presynaptic KA-Rs at glutamatergic synapses may actually have a facilitatory effect on excitatory synaptic transmission (Bortolotto et al. 1999; Chittajallu et al. 1996; Contractor et al. 2000, 2001; Lauri et al. 2001b; Schmitz et al. 2001; see Huettner 2001 for review). For instance, reduced mossy fiber plasticity has been demonstrated in GluR6 knockout mice suggesting that KA-Rs act as presynaptic autoreceptors to facilitate synaptic transmission at these synapses (Contractor et al. 2001). In addition, exogenous activation of KA-Rs appears to have a biphasic effect on synaptic transmission at both the CA3-CA1 synapse (Chittajallu et al. 1996) and the mossy fiber synapse (Schmitz et al. 2001); where low concentrations of KA (<300 nM) facilitate, and high concentrations of KA (500-5,000 nM) inhibit neurotransmitter release (Schmitz et al. 2001). It has also been demonstrated that the activation of presynaptic KA-Rs contributes to the frequency facilitation observed at the mossy fiber synapse (Lauri et al. 2001b; Schmitz et al. 2001). Furthermore, Lauri et al. (2001a), suggest that facilitatory presynaptic KA-Rs are involved in long-term potentiation observed at mossy fiber synapses. In the current study, EPSCs evoked by single shock stimulation were always inhibited by KA-R activation, while facilitation of NMDA EPSCs was occasionally observed in the PPF experiment, where Peak 2 was transiently facilitated in a subset of cells. This observation might suggest that KA-R-mediated facilitation of EPSCs in NAcc neurons may be detected with lower concentrations of KA (<50 nM) or with different stimulation protocols. Thus the physiological role of presynaptic KA-Rs in the NAcc may be dependent on the magnitude and/or activation frequency of glutamatergic inputs.

A variety of models have been used to describe the possible mechanism(s) underlying the modulation of synaptic transmission by KA (for review see Ben-Ari and Cossart 2000; Frerking and Nicoll 2000). In the hippocampus, a number of studies have demonstrated direct presynaptic ionotropic (Schmitz et al. 2001) and metabotropic (Frerking et al. 2001; Rodriguez-Moreno and Lerma 1998) actions for KA-Rs. Studies at a number of different synapses in the hippocampus, as well as in other brain regions, suggest that the presynaptic effects of KA are mediated by the indirect activation of other receptor systems (Chergui et al. 2000; Frerking et al. 1999). For instance, KA could act at somatodendritic KA-Rs on local neurons to cause the release of neuromodulators, such as adenosine or GABA, which then act heterosynaptically on the terminals of glutamatergic synapses. Similar indirect mechanisms have been shown to mediate at least part of the apparent presynaptic effects of KA at inhibitory synapses in both the hippocampus (Frerking et al. 1999) and the dorsal striatum (Chergui et al. 2000). The mGluR system has also been hypothesized to underlie KA modulation of neurotransmitter release (Schmitz et al. 2001). The current study demonstrates that the inhibition of glutamatergic synaptic transmission by KA in the NAcc does not require the secondary activation of mGluRs, GABA_B receptors, or adenosine receptors. While these data support a direct presynaptic action for KA-Rs at glutamatergic synapses in the NAcc, it should be noted that the indirect activation of other receptor systems present in the ventral striatal slice could mediate the apparent presynaptic effect of KA in this brain region.

In the current study we observed no significant intercell variability in the response of EPSCs to KA-R activation. This result might suggest that there is a homogenous population of presynaptic KA-Rs present at all glutamatergic synapses in the NAcc. However, specific glutamatergic afferents in the coronal ventral striatal slice were not discretely activated in this study. Therefore, although many of the brain regions that provide glutamatergic innervation to the NAcc express KA-R subunits (Janssens and Lesage 2001; Meador-Woodruff et al. 2001), it is possible that these different excitatory inputs do not all possess presynaptic KA-Rs or contain presynaptic KA-Rs with different subunit compositions.

In summary, there is growing evidence that KA-Rs play an integral role in regulating synaptic activity in many brain regions (Ben-Ari and Cossart 2000; Frerking and Nicoll 2000). The current study suggests that functional KA-Rs are also present in the NAcc and that activation of these receptors may have both pre- and postsynaptic sequelae. Elucidation of the overall physiological role of KA-Rs in the NAcc will require further study. It will also be important to evaluate the possible involvement of these KA-Rs in the pathophysiology of disorders associated with perturbations in NAcc synaptic activity such as schizophrenia and drug and alcohol addiction.