Membrane properties and synaptic connectivity of fast-spiking interneurons in rat ventral striatum

BR A IN RE S E A RCH 1 1 52 ( 20 0 7 ) 4 9 –5 6

a v a i l a b l e a t w w w. s c i e n c e d i r e c t . c o m

w w w. e l s e v i e r. c o m / l o c a t e / b r a i n r e s

Research Report

Membrane properties and synaptic connectivity of fast-spiking interneurons in rat ventral striatum Stefano Taverna a,b , Barbara Canciani b , Cyriel M.A. Pennartz a,c,⁎ a

Graduate School Neurosciences Amsterdam, The Netherlands Vrije Universiteit Medical Center, Department of Anatomy, P.O. Box 7057, MF-G102, 1007 MC, Amsterdam, The Netherlands c University of Amsterdam, Swammerdam Institute for Life Sciences, Center for Neuroscience, Department of Animal Physiology and Cognitive Neuroscience, Amsterdam, The Netherlands b

A R T I C LE I N FO

AB S T R A C T

Article history:

In vitro patch-clamp recordings were made to study the membrane properties and synaptic

Accepted 11 March 2007

connectivity of fast-spiking interneurons in rat ventral striatum. Using a whole-cell

Available online 24 March 2007

configuration in acutely prepared slices, fast-spiking interneurons were recognized based

Keywords:

immunocytochemistry. Membrane properties of fast-spiking interneurons were

Bursting

distinguished from those of medium-sized spiny neurons by their more depolarized

Fast-spiking

resting membrane potential, lower action potential amplitude and shorter half-width, short

Ventral striatum

spike repolarization time and deep spike afterhyperpolarization. Firing patterns of

Interneurons

interneurons could be subdivided in a bursting and non-bursting mode. Simultaneous

Nucleus accumbens

dual whole-cell recordings revealed a high degree of connectivity of fast-spiking

Parvalbumin

interneurons to medium-sized spiny neurons via unidirectional synapses. Burst firing in

Patch clamp

fast-spiking interneurons that were presynaptic to medium-sized spiny neurons resulted in

on their firing properties and their morphological phenotype was confirmed by

barrages of postsynaptic potentials showing an initial amplitude increment, rapidly followed by a decrement. In conclusion, ventral striatal fast-spiking interneurons can be clearly distinguished from medium-sized spiny neurons by their membrane properties and their firing patterns can be subdivided in bursting and non-bursting modes. Their synaptic connectivity to medium-sized spiny neurons is unidirectional and characterized by frequency-dependent, dynamic changes in postsynaptic amplitude. © 2007 Elsevier B.V. All rights reserved.

1.

Introduction

Fast-spiking interneurons (FSIs) constitute a class of interneurons of the ventral striatum (VS), a component of the striatum that plays a role in expressing and adjusting goaldirected and emotional behaviour (Mogenson et al., 1980;

Pennartz et al., 1994; Cardinal et al., 2002; Carelli, 2002; Nicola et al., 2004a,b). FSIs contain aspiny dendrites, emit axon collaterals that spread locally but may also reach relatively distant subregions of the striatum, and express the calciumbuffering protein parvalbumin (PV; Kawaguchi 1993; Kawaguchi et al., 1995). Striatal FSIs are thought to exert inhibitory

⁎ Corresponding author. Graduate School of Neurosciences Amsterdam, University of Amsterdam, Faculty of Science, Swammerdam Institute for Life Sciences, P.O. Box 94084, Kruislaan 320, 1090 GB, Amsterdam, The Netherlands. Fax: +31 20 525 7709. E-mail address: [email protected] (C.M.A. Pennartz). Abbreviations: AHP, afterhyperpolarization; dPSP, depolarizing postsynaptic potential; FSI, fast-spiking interneuron; MSN, mediumsized spiny neuron; NGS, normal goat serum; PV, parvalbumin; TBS, Tris buffer solution; VS, ventral striatum 0006-8993/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.brainres.2007.03.053

50

BR A IN RE S EA RCH 1 1 52 ( 20 0 7 ) 4 9 –56

control over the excitability of the principal cells of the striatum, i.e. medium-sized spiny neurons (MSNs; Kita et al., 1990; Pennartz and Kitai, 1991; Kita, 1996; Plenz and Kitai, 1998; Koos and Tepper, 1999; Koos et al., 2004). Patch-clamp studies in the dorsal striatum have shown that FSIs are characterized by ‘fast’ (i.e. short-lasting) action potentials discharged at high maximal rates (∼ 200 Hz), prominent frequency accommodation when moderately depolarized, a relatively large spike afterhyperpolarization (AHP) and subthreshold oscillations

occurring at a frequency of ∼40 Hz (Kawaguchi, 1993; Koos and Tepper, 1999). Dual-cell recordings in cultured (Plenz and Kitai, 1998) and acutely prepared striatal slices (Koos and Tepper, 1999, 2002; Koos et al., 2004) presented evidence for GABAA receptor mediated inhibition by FSIs onto MSNs. In vitro, this inhibition is expressed by way of blocking or delaying postsynaptic firing of action potentials (Koos and Tepper, 1999). An additional type of interaction, which has been indicated by recordings in hippocampus but has not been

Fig. 1 – Electrophysiological and anatomical properties of medium-sized spiny neurons (MSNs) and fast-spiking interneurons (FSIs). (A, B) Top, examples of current clamp recordings of MSN (left) and FSI (right) firing patterns in response to injection of a current pulse (100 pA, 2 s). Resting membrane potentials were −78 mV for MSN and −63 mV for FSI. Inset: magnification of subthreshold membrane potential oscillations elicited by current injection in FSIs. Bottom, current–voltage plots pertaining to a MSN (left) and FSI (right). Inward rectification at negative voltage levels is visible as a deviation from the linearly extrapolated portion of the curve (dotted line). (C) Example of a MSN filled with biocytin via the recording pipette and stained with DAB-nickel. Note the presence of dendritic spines. (D) FSI visualized by incubation with anti-parvalbumin antibody and staining with DAB-nickel. Four primary non-spiny dendrites are visible. (E) Fluorescent image of FSIs previously incubated with anti-PV antibody and streptavidin–Alexa 594 conjugate, (F) Double staining for biocytin and PV. Both FSIs and MSNs were positive to biocytin staining, whereas only FSIs were also positive to PV.

51

BR A IN RE S E A RCH 1 1 52 ( 20 0 7 ) 4 9 –5 6

documented for the striatum, is that an initial inhibition of principal cells by FSIs is followed by a rebound excitation (Buhl et al., 1995; Cobb et al., 1995). In view of anatomical and electrophysiological evidence, it has been suggested that FSIs in the VS provide feed-forward inhibition of MSNs, primarily by rapid shunting of glutamatergic limbic and prefrontal inputs and thus imposing a temporal window for suppressing postsynaptic excitability and synaptic plasticity in MSNs (Pennartz and Kitai, 1991; Pennartz et al., 1993, 1994). It remains unknown, however, whether the membrane properties and connectivity of FSIs in the VS are similar to those in the dorsal striatum. A separate investigation of FSIs in the VS is warranted considering recently presented functional differences in entrainment of FSI firing to EEG rhythms in the dorsal and ventral striatum in vivo (Berke et al., 2004). The present paper describes the basic membrane properties of FSIs in the VS and their synaptic connections to medium-sized spiny neurons. In particular, the current findings highlight the bursting capacity of at least a subgroup of FSIs as well as a dynamic, frequency-dependent effect in the amplitude of synaptic potentials observed in MSNs following presynaptic burst firing of FSIs.

2.

Results

2.1.

Single-cell recordings: membrane properties

Current-clamp recordings were made from a total of 137 cells, mainly located in the VS core. Rectangular current pulses (−200 to +300 pA) were injected via the patch electrode to study their membrane properties. Of these cells, 122 (89%) showed characteristics of MSNs, i.e. slightly or non-adapting spike trains with peak firing rates reaching 30 Hz, inward rectification at hyperpolarized potentials, and subthreshold ramp-like depolarizations (Fig. 1A). Fifteen cells (11%) fired action potentials at higher rates (up to 150 Hz) with strong frequency adaptation, also displayed inward rectification upon hyperpolarization, and showed fast subthreshold oscillations at depolarized levels of membrane potential (Fig. 1B). These properties are typical of FSIs as described earlier in the dorsal striatum (Kawaguchi, 1993; Kita, 1993; Koos and Tepper, 1999). In comparison to MSNs, FSIs had a significantly less negative resting membrane potential, lower spike amplitude (measured with respect to firing threshold) but higher spike AHP amplitude, a shorter action potential half-width and a roughly tenfold shorter action potential repolarization time (Table 1). Out of 15 FSIs, 10 (67%) responded to suprathreshold injection of current steps (duration: 1 s) by firing a train of 2 to 15 action potentials with a short latency (10–50 ms) from pulse onset, displaying strong frequency adaptation after 50–300 ms. In this subgroup, interspike intervals within the trains were highly variable and no marked periodicity was found (Fig. 1B). In contrast, 5 FSIs (33%) responded to suprathreshold current stimuli by firing bursts of 2 up to 16 action potentials (Fig. 2). During a 1- or 2-s lasting pulse, some adaptation of the interburst frequency occurred (Fig. 2A). When burst activity was evoked by steady suprathreshold current injection for 10–40 s, inter-burst intervals of different lengths were observed (Fig. 2B). The average inter- and intra-burst frequencies across the

Table 1 – Electrophysiological properties of MSNs and FSIs

Rmp (mV) Rin (MΩ) Ap amplitude (mV) Ap threshold (mV) Ap half-width (ms) dV/dt of Ap rising slope phase (V/s) Ap repolarization time (ms) AHP (mV)

MSN (n = 15)

FSI (n = 15)

−76 ± 5 207 ± 21 50 ± 4 −36 ± 1 1.6 ± 0.1 109 ± 9 33 ± 5 6.3 ± 0.3

− 67 ± 5* 226 ± 24 38 ± 2** − 33 ± 1 1.0 ± 0.1* 93 ± 9 3 ± 0.2** 11.6 ± 1.1*

Values are mean ± S.E.M. Abbreviations: R mp : resting membrane potential, R in : input resistance, Ap: action potential, AHP: afterhyperpolarizing potential. The AP amplitude was measured with respect to firing threshold level. Ap repolarization time refers to the time window between the spike peak and the AHP minimum value. *p < 0.05, **p < 0.005; Mann–Whitney's U-test.

5 cells were 2.4 ± 0.9 Hz (range 0.6–5.6) and 52 ± 13 Hz (range 20– 80), respectively. Conversely, firing activity of non-bursting FSIs occurred across a continuum of variable spike intervals (Fig. 2C). Neither non-bursting nor bursting FSIs showed spontaneous firing activity. Spontaneous alternation between upand down-states in membrane potential was not detected. Finally, membrane physiological and anatomical properties (resting membrane potential, input resistance, area of soma) of bursting and non-bursting FSIs did not differ significantly.

2.2.

Morphology and immunocytochemistry

MSNs and FSIs presented morphological differences revealed by immunohistochemical staining (Fig. 1). All of the stained MSNs (n = 10) were PV-negative and were characterized by densely spined dendrites (Figs. 1C, F). Conversely, FSIs were labeled by anti-PV antibodies (n = 9 out of 10) and all had aspiny and varicose dendrites (Figs. 1D, E, F). For unknown reasons, PV-immunoreactivity could not be confirmed for one neuron that was electrophysiologically classified as FSI. Both cell types had fusiform, polygonal, or round (spheroid) somata. Average somatic areas of FSIs were significantly larger than those of MSNs (150 ± 15 μm2 and 63 ± 4 μm2, respectively, p < 0.005, Mann–Whitney's U-test). These data confirm the phenotypical differences between FSIs and MSNs described previously in the dorsal striatum (Bolam et al., 1985; Pasik et al., 1988; Cowan et al., 1990; Kita, 1993).

2.3.

Dual-cell recordings: synaptic connectivity

We recorded pairs of cells consisting of one FSI interneuron and one MSN (Fig. 3). Three out of five pairs (60%) of this nature were synaptically connected. In all three cases the FSIs were of the bursting type. Action potentials evoked by suprathreshold stimulation of the FSI induced a synaptic response in the MSN, whereas stimulation of the MSN failed to induce any response in the FSI. Bursts of depolarizing postsynaptic potentials (dPSPs) were time-locked to bursts of presynaptic action potentials (Fig. 3). It should be noted that the depolarizing nature of these postsynaptic potentials directly relates to the

52

BR A IN RE S EA RCH 1 1 52 ( 20 0 7 ) 4 9 –56

depolarized reversal potential for Cl− ions (−20.8 mV) imposed by the composition of the pipette solution, which was chosen to enhance detectability of postsynaptic responses. Thus, their depolarizing nature does not indicate any excitatory function of FSIs. Current pulses were applied to the presynaptic cell to elicit approximately passive depolarization, as noted from the absence of spikes, depolarizing ramps or other manifestations of intrinsic voltage-dependent membrane currents. These pulses did not elicit any visible response in the postsynaptic cell, arguing against the presence of electrical synapses. In most barrages of dPSPs (∼ 90%), we observed temporal summation of the first 2–5 dPSPs followed by depression of the remaining 2–5 dPSPs (Fig. 3B). In a minority of dPSP

Fig. 3 – Bursting FSIs and MSNs are synaptically connected. (A) Depolarizing postsynaptic potentials (dPSPs) elicited in a MSN by firing of action potentials of a bursting FSI (B). The portion of the traces included in the dotted frame is enlarged in the lower two panels of B. Note the initial frequencyfacilitation and subsequent depression of dPSPs in response to presynaptic action potential bursts.

Fig. 2 – Bursting activity in FSIs. (A) Injection of a suprathreshold current pulse (100 pA, 1 s) induces burst-like firing in a FSI. (B) Bursting activity is maintained over relatively long periods of steady current injection (200 pA, 10 s). (C) Comparison between irregular, non-bursting firing (top) and bursting firing (bottom) in two different FSIs. Autocorrelation plots were created for each of the two cells (right; bin size: 1 ms). Dotted lines represent threshold levels of significance (equal to 3*SD value, assuming a Poisson distribution of the spike count). The irregular-firing cell (top) shows non-significant, evenly distributed peaks throughout a ±200 ms time window, whereas the burst-firing cell (bottom) is characterized by several pronounced peaks crossing the level of significance and symmetrically placed around 0 ms. Note that the maximum number of spikes at 0 ms is not shown in the lower graph.

barrages (10%), mostly derived from the same cell pair, the first evoked dPSP had a relatively high amplitude (4.0 ± 0.3 mV) and was followed by dPSPs of lower amplitude. The onset latency of the first dPSP following an action potential was 1.1 ± 0.1 ms (n = 3 pairs, 56 dPSPs). The average amplitude and 10–90% rise time of the first dPSP in a burst were 1.5 ± 1.1 mV and 5 ± 1 ms, respectively. The decay time constant of the last dPSP in a burst was 40 ± 10 ms. Transmission failures were nearly absent in two out of three pairs tested with a total of 40 presynaptic spikes, whereas failures in the third pair occurred at a rate of 29% (n = 16 presynaptic spikes).

3.

Discussion

3.1.

Membrane properties of fast spiking interneurons

In this paper we describe basic membrane properties, firing patterns and synaptic connectivity of FSIs of the rat VS, as investigated by single and dual-cell patch-clamp recordings in acutely prepared slices. FSIs recorded in whole-cell mode in slices were characterized, amongst others, by relatively high maximal firing rates (up to 150 Hz), frequency adaptation in

BR A IN RE S E A RCH 1 1 52 ( 20 0 7 ) 4 9 –5 6

evoked spike trains, fast spike repolarization and prominent subthreshold oscillations in membrane potential (Table 1). Several membrane properties of FSIs provided clear physiological markers for distinguishing them from medium-sized spiny neurons, viz. their more depolarized resting membrane potential, lower action potential amplitude and shorter halfwidth, short spike repolarization time and deep spike afterhyperpolarization. Of these markers, the tenfold lower spike repolarization time is particularly striking. Post-fixation staining revealed anatomical characteristics typical of FSIs such as relatively large somata (average area ∼ 150 μm2), immunoreactivity to PV and absence of dendritic spines. These properties of VS FSIs appear similar to those reported for the dorsal striatum (cf. Kawaguchi, 1993; Kawaguchi et al., 1995; Plenz and Kitai, 1998; Cowan et al., 1990; Kita, 1993; Bennett and Bolam, 1994; Ramanathan et al., 2002). A subpopulation of FSIs (33%) fired bursts of action potentials when stimulated overthreshold (Fig. 2). Preliminary data (not shown) indicate that an individual FSI, depending on its input, may switch between a bursting and non-bursting mode of firing, suggesting that bursting is not an exclusive property of a fixed subclass of FSI cells, but can be elicited from cells that do not appear to be bursting at first glance. However, more research will be needed to investigate which type of inputs may systematically elicit switching between burst and non-burst firing, and whether FSIs can fire bistably, i.e., in these two firing modes only, or show gradual transitions between them. Moreover, further work is needed to examine whether burst firing of FSIs can be distinguished in vivo against a high background firing rate, either during awake or various sleep conditions.

3.2.

Circuit connections of fast spiking interneurons

In three FSIs that were synaptically connected to MSNs (Fig. 3), presynaptic bursts of action potentials gave rise to barrages of dPSPs in MSNs that were first frequency-facilitated but subsequently showed a depression. The initial frequencyfacilitation is likely due, at least in part, to a temporal summation effect. Since slices were perfused with ACSF containing the AMPA-receptor antagonist NBQX and striatal FSIs are known to use GABA as neurotransmitter (Cowan et al., 1990; Kawaguchi, 1993), dPSPs recorded in response to FSI stimulation were most likely due to GABAA-receptor activation (Koos and Tepper, 1999; Koos et al., 2004). These data, albeit limited in number, suggest that FSIs have a powerful, frequency-facilitated effect on MSNs in the VS.

3.3.

Functional implications

Slow EEG rhythms at 0.5–3 Hz (delta frequency) have been recorded in the VS of freely moving rats during awake immobility and face washing (Leung and Yim, 1993). A delta rhythm was also correlated with membrane potential oscillations of cells recorded intracellularly during anaesthesia. Furthermore, local field potential oscillations at theta (∼ 8 Hz), beta (∼ 20 Hz) and gamma (∼50 Hz) frequencies have been recorded in the VS of freely moving rats during behavioral performance, but not during slow-wave and REM sleep (Berke et al., 2004).

53

These delta and gamma frequencies match the intrinsic bursting activity of FSIs we show here (burst frequency: 2.4 ± 0.9 Hz; intraburst firing rate: 52 ± 13 Hz). Monosynaptic connectivity between bursting FSIs and MSNs (Fig. 3) provides a potential pathway by which FSIs may phase MSN firing so as to engage in slow rhythms under particular behavioural conditions such as face washing. We did not find evidence for GABAergic projections from MSNs to FSIs, arguing against the existence of abundant recurrent inhibition. Although we did not directly assess the possibility that FSIs drive MSNs firing into a particular rhythm, barrages of dPSPs in MSNs elicited by firing bursts of FSIs may sculpt postsynaptic firing activity to occur at certain frequencies. Importantly, the first 2–5 dPSPs in a postsynaptic barrage showed prominent frequency-facilitation (Fig. 3). This may help entrain MSN firing activity to a specific rhythm, because temporal summation resulting from high-frequency, burst-like inputs may override effects of asynchronous, non-summating inputs (Singer and Gray, 1995). The presence of a powerful, frequency-facilitated FSI-to-MSN synaptic response also agrees with the previously postulated existence of a feed-forward inhibitory circuit in the ventral (Pennartz and Kitai, 1991) as well as dorsal striatum (Kita, 1993; Mallet et al., 2005). Although it remains to be determined which anatomical sources of glutamatergic afferents synapse directly onto FSIs in the VS, it is reasonable to hypothesize that at least hippocampal and prefrontal outputs facilitate firing of FSIs and thereby evoke a shunting inhibition in MSNs (Pennartz and Kitai, 1991; Pennartz et al., 1993). The function of FSImediated feedforward inhibition may not only be to curtail the impact of incoming excitatory waves at the MSN soma and resulting spikes, but also to control synaptic plasticity, e.g. induction of long-term potentiation, which is under GABAergic control in the VS (Pennartz et al., 1993). It will be important in the future to perform dual-cell recordings using a physiological Cl− reversal potential, and directly study the impact of bursting activity of FSIs on firing patterns of postsynaptic MSNs. The idea that electrotonically coupled striatal FSIs (cf. Koos and Tepper, 1999) might serve to pace and synchronize MSNs to massive intrastriatal EEG activity is consistent with studies in cortical areas, where electrical coupling between interneurons has been demonstrated as well (cf. Buzsaki and Chrobak, 1995; Whittington and Traub, 2003). Neocortical and hippocampal GABAergic interneurons have been shown to be involved in the generation of coherent oscillatory activity of large populations of cells, mainly in the theta and gamma frequency ranges (Cobb et al., 1995; Tamas et al., 2000; Whittington and Traub, 2003; Freund, 2003). However, in contrast to hippocampal studies reporting a post-inhibitory rebound excitation in pyramidal cells (Buhl et al., 1995; Cobb et al., 1995), we have failed to observe such an effect in MSNs of the VS, at least in vivo (data not shown). In that sense, FSI functions observed in cortical structures should not be assumed to hold in subcortical areas such as the VS. A slow pacemaking activity (0.4 to 1.8 Hz) has been described in cultured slices of the subthalamic nucleus – external globus pallidus network (Plenz and Kitai, 1999). About 25% of pallidal neurons give rise to collateral projections to the striatum (Kita, 1993; Bevan et al., 1998; Bolam et al., 2000) and

54

BR A IN RE S EA RCH 1 1 52 ( 20 0 7 ) 4 9 –56

the majority of the pallidostriatal terminals selectively innervates FSIs (Bevan et al., 1998). Reciprocal connectivity between the VS and the ventral pallidum has been indicated (Hakan et al., 1992; Groenewegen et al., 1993). Thus, it is possible that the pallidostriatal pathway conveys slow pacemaking, rhythmic discharges to the striatum via FSIs, which in turn disengage MSNs via GABAergic synapses (Bolam et al., 2000). However, the nature of the synaptic inputs necessary to induce burst firing in VS FSIs is still unknown, and more work is needed to understand how excitatory and inhibitory inputs onto FSIs shape burst-like and non-bursting firing patterns. In conclusion, our data show membrane physiological properties of ventral FSIs that appear similar to those encountered in dorsal striatal FSIs and clearly distinguish them from MSNs in the VS. Of particular interest are the distinct bursting mode observed in at least a subset of ventral FSIs and a powerful and frequency-facilitated inhibition from FSIs onto MSNs, which does not appear to be reciprocated. Topics to be addressed in future research include the role of FSIs in regulating the in vivo firing behavior of striatal projection neurons, both under awake and sleeping conditions (cf. Pennartz et al., 2004), the relationship between FSI firing and intrastriatal oscillations in local field potential and the function of FSIs in fast information processing, particularly in relation to prediction and processing of positive or negative reinforcement.

4.

Experimental procedures

4.1.

Slice preparation and electrophysiology

Wistar rats (23 to 30 days of age; Harlan, Horst, The Netherlands) were anaesthetized with an intraperitoneal injection of Nembutal (60 mg/kg) and decapitated. All procedures were performed following the national guidelines for the use of vertebrate animals in research. Slices containing the VS were prepared as described previously (Taverna and Pennartz, 2003; Taverna et al., 2004). Briefly, brains were removed from the skull and 280-μm-thick coronal slices were cut in artificial cerebrospinal fluid (ACSF) at 4 °C using a vibratome (VT1000S, Leica, Germany). Following a 1-h recovery period, slices were submerged in a recording chamber in which ACSF was continuously flowing (1–2 ml/ min) at 32 °C. The composition of ACSF was (in mM): NaCl 124, KCl 3.5, NaH2PO4 1, CaCl2 2.5, NaHCO3 26, MgSO4 1.3, D-glucose 10, saturated with 95% O2, 5% O2 (pH 7.3). The AMPA-receptor antagonist NBQX (5 μM; 2,3-dioxo-6-nitro-1,2,3,4-tetrahydrobenzo [f]quinoxaline-7-sulfonamide disodium salt; Tocris Cookson, UK) was added to the ACSF at the beginning of experiments. Patch-clamp pipettes (5–7 MΩ) contained the following solution (in mM): K-gluconate 60, KCl 58, HEPES 10, EGTA 0.5, MgCl2 1, Na2-ATP 2, Na3-GTP 0.3, Na2-phosphocreatine 20, leupeptin 0.1, biocytin 26 (Sigma; solution was adjusted to pH 7.2 with KOH). We chose this composition in order to obtain a relatively depolarized reversal potential for Cl− ions (−20.8 mV at 32 °C according to the Nernst equation) as compared to the resting membrane potential. A liquid junction-potential of +3 mV between intra- and extracellular

solution was corrected for. Dual-cell recordings were made in current clamp mode using a MultiClamp 700 A amplifier (filtering at 1–3 kHz, sampling rate: 10 kHz) data were processed using pClamp 8.0 software (Molecular Devices, Sunnyvale, CA, USA). Cells were chosen under visual control using an Axioscope upright microscope (Zeiss, Germany) equipped with Hoffman modulation contrast. After wholecell configurations were established in voltage clamp (VC) mode simultaneously in two cells, the recording mode was subsequently switched to current clamp (CC) and, after bridgebalance compensation, the intrinsic membrane properties of both cells were assessed by applying positive and negative rectangular current pulses (range: +100/−200 pA, 1–2 s).

4.2.

Analysis of electrophysiological data

Pre- and postsynaptic signals recorded in current- and voltage-clamp mode were analyzed using Clampfit 8.0 software (Molecular Devices), Origin 5.0 (Microcal Software, Northampton, USA), STATISTICA (Statsoft, Tulsa, USA), and MATLAB 6.5 (The MathWorks Inc., Natick, USA). Methods to analyze intrinsic and synaptic properties of individual neurons, such as dPSP amplitude, latency and failures, were described in a previous paper (Taverna et al., 2004). We applied the following criteria to classify interneurons as ‘bursting’ or ‘non-bursting’: (i) a burst was defined as a relatively short episode (30–200 ms) in which 2–16 action potentials were fired at a regular frequency, i.e. the intra-burst interspike interval did not markedly vary across bursts (mean intra-burst coefficient of variation, equalling SD/mean, was 0.15 ± 0.02; n = 165 observations); (ii) the inter-burst frequency did not exceed 10% of the intra-burst firing rate; (iii) autocorrelograms of firing activity induced by steady current injection for a period of 20–40 s yielded multi-peaked plots in which at least two peaks crossed a significance level set as three times the standard deviation of the spike train under Poisson assumptions (corresponding to p = 0.0013; bin size: 1.0 ms; Fig. 2C). For testing of statistical differences between the electrophysiological properties of MSNs and FSIs we used Mann–Whitney's U-test. Results are given as means ± S.E.M.

4.3.

Morphology and immunocytochemistry

To examine the morphology and neurochemical phenotype of recorded neurons, slices were fixated in 4% formalin solution after terminating recordings and stored at +4 °C for a period of a few days up to a few weeks. Following cryoprotection in 30% sucrose in distilled water (1 h at room temperature, RT), slices were rapidly frozen in 30% sucrose solution onto the stage of a sliding microtome and cut into sections of 30 μm. Sections were collected in a 24-well plate containing Tris buffer solution (TBS) (0.05 M Tris–HCl supplemented with 0.15 M NaCl, pH 7.6) for direct processing. Sections were double-stained for biocytin and parvalbumin (PV) as follows: after three rinses with TBS, sections were incubated in 0.5% Triton X-100, 1% bovine serum albumin (BSA) and 5% normal goat serum (NGS) in TBS for 1 h at RT. Immediately thereafter, sections were kept in a mix of streptavidin–Alexa 594 conjugate (1:200, Molecular Probes Inc.) and anti-PV antibody (monoclonal 1:2000, Sigma, St.

BR A IN RE S E A RCH 1 1 52 ( 20 0 7 ) 4 9 –5 6

Louis, MO) in 0.1% Triton X-100, 1% BSA and 5% NGS in TBS (TBSTx–BSA–NGS) for 4 successive days at 4 °C. Following three rinses in 0.1% Triton X-100 in TBS (TBS-Tx) of 10 min each, the sections were incubated with goat anti-mouse Alexa 488 conjugate (1:200 Molecular Probes Inc.) in TBSTx–BSA–NGS for 1 h at RT. The incubation was stopped by three rinses with Tris–HCl (pH 8.0) for 10 min each. Sections were mounted on glass slides from Tris–HCl solution, containing 0.2% gelatin (Oxoid, Basingstoke, UK) and coverslipped in toluene–DPX mounting medium. To control for antibody specificity, some control sections (i.e. non-recorded) were stained to reveal PV-positive neurons. Sections were mounted on glass slides and air-dried. Mounted sections were dehydrated through an ascending series of alcohol (2× 96%, 2× 100%), left in xylene for 20 min, and finally coverslipped using Entellan. We also studied the morphology of a number of biocytinfilled cells without immunocytochemical staining for PV. For these cases, sections were rinsed three times with TBS and then incubated in 0.5% Triton X-100, 5% NGS in TBS for 1 h at RT. The sections were incubated in avidin–biotin– peroxidase complex (Vector, Burlingame, CA) in TBS-Tx for four successive days at 4 °C. Sections were stained with DAB-Ni (7.5 mg 3,3-diaminobenzidinetetrahydrochloride from Sigma, St. Louis, MO, 0.225 g nickel-ammonium sulphate, Boom Meppel, NL, 10 μl of 30% H2O2) in 50 ml Tris–HCl, pH 8.0, for 5–20 min. The DAB-Ni-stained neurons were photographed using a bright-field microscope (400×; Leica, Germany). Fluorescence and confocal laser scanning microscopes (Leica, Germany) were used to examine cell morphology and immunocytochemical phenotype. Somatic areas of neurons were drawn and analyzed by Neurolucida MicroBrightfield Inc. at 400×.

Acknowledgments We are grateful to Henk J. Groenewegen for his advice and making the materials for immunocytochemistry available. This work was supported by The Netherlands Organisation for Scientific Research Grant 903-47-092, HFSPO grant RGP-0127 and Grant BSIK-03053 from SenterNovem, The Netherlands.

REFERENCES

Bennett, B.D., Bolam, J.P., 1994. Synaptic input and output of parvalbumin-immunoreactive neurons in the neostriatum of the rat. Neuroscience 62, 707–719. Berke, J.D., Okatan, M., Skurski, J., Eichenbaum, H.B., 2004. Oscillatory entrainment of striatal neurons in freely moving rats. Neuron 43, 883–896. Bevan, M.D., Booth, P.A., Eaton, S.A., Bolam, J.P., 1998. Selective innervation of neostriatal interneurons by a subclass of neuron in the globus pallidus of the rat. J. Neurosci. 18, 9438–9452. Bolam, J.P., Powell, J.F., Wu, J.Y., Smith, A.D., 1985. Glutamate decarboxylase-immunoreactive structures in the rat neostriatum: a correlated light and electron microscopic study including a combination of Golgi impregnation with immunocytochemistry. J. Comp. Neurol. 237, 1–20.

55

Bolam, J.P., Hanley, J.J., Booth, P.A., Bevan, M.D., 2000. Synaptic organisation of the basal ganglia. J. Anat. 196, 527–542. Buhl, E.H., Cobb, S.R., Halasy, K., Somogyi, P., 1995. Properties of unitary IPSPs evoked by anatomically identified basket cells in the rat hippocampus. Eur. J. Neurosci. 7, 1989–2004. Buzsaki, G., Chrobak, J.J., 1995. Temporal structure in spatially organized neuronal ensembles: a role for interneuronal networks. Curr. Opin. Neurobiol. 5, 504–510. Cardinal, R.N., Parkinson, J.A., Hall, J., Everitt, B.J., 2002. Emotion and motivation: the role of the amygdala, ventral striatum, and prefrontal cortex. Neurosci. Biobehav. Rev. 26, 321–352. Carelli, R.M., 2002. Nucleus accumbens cell firing during goal-directed behaviors for cocaine vs. ‘natural’ reinforcement. Physiol. Behav. 76, 379–387. Cobb, S.R., Buhl, E.H., Halasy, K., Paulsen, O., Somogyi, P., 1995. Synchronization of neuronal activity in hippocampus by individual GABAergic interneurons. Nature 378, 75–78. Cowan, R.L., Wilson, C.J., Emson, P.C., Heizmann, C.W., 1990. Parvalbumin-containing GABAergic interneurons in the rat neostriatum. J. Comp. Neurol. 302, 197–205. Freund, T.F., 2003. Interneuron diversity series: rhythm and mood in perisomatic inhibition. Trends Neurosci. 26, 489–495. Groenewegen, H.J., Berendse, H.W., Haber, S.N., 1993. Organization of the output of the ventral striatopallidal system in the rat: ventral pallidal efferents. Neuroscience 57, 113–142. Hakan, R.L., Berg, G.I., Henriksen, S.J., 1992. Electrophysiological evidence for reciprocal connectivity between the nucleus accumbens septi and ventral pallidal region. Brain Res. 581, 344–350. Kawaguchi, Y., 1993. Physiological, morphological, and histochemical characterization of three classes of interneurons in rat neostriatum. J. Neurosci. 13, 4908–4923. Kawaguchi, Y., Wilson, C.J., Augood, S.J., Emson, P.C., 1995. Striatal interneurones: chemical, physiological and morphological characterization. Trends Neurosci. 18, 527–535. Kita, H., 1993. GABAergic circuits of the striatum. Prog. Brain Res. 99, 51–72. Kita, H., 1996. Glutamatergic and GABAergic postsynaptic responses of striatal spiny neurons to intrastriatal and cortical stimulation recorded in slice preparations. Neuroscience 70, 925–940. Kita, H., Kosaka, T., Heizmann, C.W., 1990. Parvalbuminimmunoreactive neurons in the rat neostriatum: a light and electron microscopic study. Brain Res. 536, 1–15. Koos, T., Tepper, J.M., 1999. Inhibitory control of neostriatal projection neurons by GABAergic interneurons. Nat. Neurosci. 2, 467–472. Koos, T., Tepper, J.M., 2002. Dual cholinergic control of fast-spiking interneurons in the neostriatum. J. Neurosci. 22, 529–535. Koos, T., Tepper, J.M., Wilson, C.J., 2004. Comparison of IPSCs evoked by spiny and fast-spiking neurons in the neostriatum. J. Neurosci. 24, 7916–7922. Leung, L.S., Yim, C.Y., 1993. Rhythmic delta-frequency activities in the nucleus accumbens of anesthetized and freely moving rats. Can. J. Physiol. Pharmacol. 71, 311–320. Mallet, N., Le Moine, C., Charpier, S., Gonon, F., 2005. Feedforward inhibition of projection neurons by fast-spiking GABA interneurons in the rat striatum in vivo. J. Neurosci. 3857–3869. Mogenson, G.J., Jones, D.L., Yim, C.Y., 1980. From motivation to action: functional interface between the limbic system and the motor system. Prog. Neurobiol. 14, 69–97. Nicola, S.M., Yun, I.A., Wakabayashi, K.T., Fields, H.L., 2004a. Cue-evoked firing of nucleus accumbens neurons encodes motivational significance during a discriminative stimulus task. J. Neurophysiol. 91, 1840–1865. Nicola, S.M., Yun, I.A., Wakabayashi, K.T., Fields, H.L., 2004b. Firing of nucleus accumbens neurons during the consummatory phase of a discriminative stimulus task depends on previous reward predictive cues. J. Neurophysiol. 91, 1866–1882.

56

BR A IN RE S EA RCH 1 1 52 ( 20 0 7 ) 4 9 –56

Pasik, P., Pasik, T., Holstein, G.R., Hamori, J., 1988. GABAergic elements in the neuronal circuits of the monkey neostriatum: a light and electron microscopic immunocytochemical study. J. Comp. Neurol. 270, 157–170. Pennartz, C.M.A., Kitai, S.T., 1991. Hippocampal inputs to identified neurons in an in vitro slice preparation of the rat nucleus accumbens: evidence for feed-forward inhibition. J. Neurosci. 11, 2838–2847. Pennartz, C.M.A., Ameerun, R.F., Lopes da Silva, F.H., 1993. Synaptic plasticity in an in vitro slice preparation of the rat nucleus accumbens. Eur. J. Neurosci. 5, 107–117. Pennartz, C.M.A., Groenewegen, H.J., Lopes da Silva, F.H., 1994. The nucleus accumbens as a complex of functionally distinct neuronal ensembles: an integration of behavioural, electrophysiological and anatomical data. Prog. Neurobiol. 42, 719–761. Pennartz, C.M.A., Lee, E., Verheul, J., Lipa, P., Barnes, C.A., McNaughton, B.L., 2004. The ventral striatum in off-line processing: ensemble reactivation during sleep and modulation by hippocampal ripples. J. Neurosci. 24, 6446–6456. Plenz, D., Kitai, S.T., 1998. Up and down states in striatal medium spiny neurons simultaneously recorded with spontaneous activity in fast-spiking interneurons studied in cortex–striatum–substantia nigra organotypic cultures. J. Neurosci. 18, 266–283.

Plenz, D., Kitai, S.T., 1999. A basal ganglia pacemaker formed by the subthalamic nucleus and external globus pallidus. Nature 400, 677–682. Ramanathan, S., Hanley, J.J., Deniau, J.M., Bolam, J.P., 2002. Synaptic convergence of motor and somatosensory cortical afferents onto GABAergic interneurons in the rat striatum. J. Neurosci. 22, 158–169. Singer, W., Gray, C.M., 1995. Visual feature integration and the temporal correlation hypothesis. Annu. Rev. Neurosci. 18, 555–586. Tamas, G., Buhl, E.H., Lorincz, A., Somogyi, P., 2000. Proximally targeted GABAergic synapses and gap junctions synchronize cortical interneurons. Nat. Neurosci. 3, 366–371. Taverna, S., Pennartz, C.M.A., 2003. Postsynaptic modulation of AMPA- and NMDA-receptor currents by Group III metabotropic glutamate receptors in rat nucleus accumbens. Brain Res. 976, 60–68. Taverna, S., Dongen van, Y.C., Groenewegen, H.J., Pennartz, C. M.A., 2004. Direct physiological evidence for synaptic connectivity between medium-sized spiny neurons in rat nucleus accumbens in situ. J. Neurophysiol. 91, 1111–1121. Whittington, M.A., Traub, R.D., 2003. Interneuron diversity series: inhibitory interneurons and network oscillations in vitro. Trends Neurosci. 26, 676–682.