JN Fuel your research with LabChart
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

J Neurophysiol 96: 2295-2306, 2006. First published July 26, 2006; doi:10.1152/jn.01040.2005
0022-3077/06 $8.00
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
96/5/2295    most recent
01040.2005v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via ISI Web of Science (6)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Gale, S. D.
Right arrow Articles by Perkel, D. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Gale, S. D.
Right arrow Articles by Perkel, D. J.

Physiological Properties of Zebra Finch Ventral Tegmental Area and Substantia Nigra Pars Compacta Neurons

Samuel D. Gale1 and David J. Perkel2

1Graduate Program in Neurobiology and Behavior and 2Departments of Biology and Otolaryngology, University of Washington, Seattle, Washington

Submitted 3 October 2005; accepted in final form 18 July 2006


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 GRANTS
 ACKNOWLEDGMENTS
 REFERENCES
 
The neurotransmitter dopamine plays important roles in motor control, learning, and motivation in mammals and probably other animals as well. The strong dopaminergic projection to striatal regions and more moderate dopaminergic projections to other regions of the telencephalon predominantly arise from midbrain dopaminergic neurons in the substantia nigra pars compacta (SNc) and ventral tegmental area (VTA). Homologous dopaminergic cell groups in songbirds project anatomically in a manner that may allow dopamine to influence song learning or song production. The electrophysiological properties of SNc and VTA neurons have not previously been studied in birds. Here we used whole cell recordings in brain slices in combination with tyrosine-hydroxylase immunolabeling as a marker of dopaminergic neurons to determine electrophysiological and pharmacological properties of dopaminergic and nondopaminergic neurons in the zebra finch SNc and VTA. Our results show that zebra finch dopaminergic neurons possess physiological properties very similar to those of mammalian dopaminergic neurons, including broad action potentials, calcium- and apamin-sensitive membrane-potential oscillations underlying pacemaker firing, powerful spike-frequency adaptation, and autoinhibition via D2 dopamine receptors. Moreover, the zebra finch SNc and VTA also contain nondopaminergic neurons with similarities (fast-firing, inhibition by the µ-opioid receptor agonist [D-Ala2, N-Me-Phe4, Gly-ol5]-enkephalin (DAMGO)) and differences (strong h-current that contributes to spontaneous firing) compared with GABAergic neurons in the mammalian SNc and VTA. Our results provide insight into the intrinsic membrane properties that regulate the activity of dopaminergic neurons in songbirds and add to strong evidence for anatomical, physiological, and functional similarities between the dopaminergic systems of mammals and birds.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
METHODS
 RESULTS
 DISCUSSION
 GRANTS
 ACKNOWLEDGMENTS
 REFERENCES
 
The substantia nigra pars compacta (SNc) and ventral tegmental area (VTA) in mammals contain dopaminergic neurons that project diffusely to several areas of the telencephalon and play crucial roles in motor control, learning, and motivation (Schultz 2002Go; Wise 2004Go). Homologous cell groups bearing the same names have been identified in the avian midbrain (Reiner et al. 2004Go). The avian SNc and VTA contain cell bodies that label positively for tyrosine-hydroxylase (TH; the rate limiting enzyme in catecholamine biosynthesis) but not dopamine-beta-hydroxylase (which is involved in converting dopamine to norepinephrine). As in mammals, these dopaminergic neurons densely innervate striatal areas of the basal ganglia and project more moderately to several other regions of the telencephalon (Durstewitz et al. 1999Go; Reiner et al. 1994Go). In addition, pharmacological agents and lesions targeting the dopaminergic system have many similar behavioral effects in birds and mammals (Durstewitz et al. 1999Go). Thus anatomical and functional features of dopaminergic systems appear to be conserved in birds and mammals.

Given the important role of dopamine in motor control and learning in mammals, it seems likely that dopamine is involved in song learning and production in songbirds. Midbrain dopaminergic neurons project to several telencephalic brain nuclei involved in learning or producing song (Appeltants et al. 2000Go, 2002Go; Bottjer 1993Go; Harding 1998Go; Lewis et al. 1981Go; Soha et al. 1996Go). Dopaminergic innervation from VTA neurons is particularly dense in a distinct region of the songbird striatum named Area X. Area X is part of a discrete basal ganglia circuit required for normal song learning and for plasticity of adult song (Bottjer et al. 1984Go; Brainard and Doupe 2000Go; Scharff and Nottebohm 1991Go; Sohrabji et al. 1990Go; Williams and Mehta 1999Go). Regulation of dopamine release and the effects of dopamine on membrane excitability, synaptic transmission, and synaptic plasticity are similar in Area X and the mammalian striatum (Ding and Perkel 2002Go, 2004Go; Ding et al. 2003Go; Gale and Perkel 2005Go), but the physiological properties and functions of dopaminergic neurons that project to Area X or other song-control nuclei are unknown. Indeed, the physiological properties of SNc and VTA neurons have not been studied in birds or any other nonmammalian species.

In mammals, SNc and VTA neurons are diverse in their physiological properties and neurotransmitters they use (see DISCUSSION). Do the avian SNc and VTA contain nondopaminergic neurons? How do the intrinsic membrane properties of avian SNc and VTA neurons contribute to regulating their activity? To understand how the SNc and VTA function in songbirds, it is imperative to have some knowledge of the diversity of cell types in these areas and their electrophysiological properties. In this study, we used in vitro whole cell, current-clamp recordings and TH immunocytochemistry to determine the electrophysiological properties of dopaminergic and nondopaminergic neurons in the SNc and VTA in zebra finches.


   METHODS
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
RESULTS
 DISCUSSION
 GRANTS
 ACKNOWLEDGMENTS
 REFERENCES
 
All procedures were approved by the University of Washington Institutional Animal Care and Use Committee. Adult (>90 days old), male zebra finches were obtained from commercial suppliers. Juvenile birds were raised in our colony.

In some birds, bilateral tracer injections were made in Area X 5–9 days before slicing. Birds were anesthetized with 40 mg/kg sodium pentobarbital administered intramuscularly. The tip of a glass pipette was filled with Alexa 488-conjugated cholera-toxin subunit B (1% dissolved in dH20; Molecular Probes, Eugene, OR), and the pipette was back-filled with 0.9% NaCl. Tracer was delivered iontophoretically using +5.2 µA (7 s on, 7 s off) for 15 min.

For slicing, birds were anesthetized with isoflurane and decapitated. The brain was removed and immersed in an ice-cold solution containing (in mM) 119 NaCl, 2.5 KCl, 1.3 MgSO4, 1 NaH2PO4, 16.2 NaHCO3, 2.5 CaCl2, 11 D-glucose, and 10 HEPES. Coronal or parasagittal slices 300–400 µm thick were cut with a vibrating microtome. Slices were stored in artificial cerebrospinal fluid (ACSF), which was made of the same components described above for the slicing solution except for replacement of HEPES with an additional 10 mM NaHCO3. The ACSF was initially ~35°C when the slices were transferred and then allowed to cool to room temperature. All solutions were continuously bubbled with a gas mixture of 95% O2-5% CO2. Slices were left for ≥1 h before use.

For recordings, a slice was transferred to a chamber where it was submerged and perfused (2–3 ml/min) with pregassed ACSF heated to 32°C. Slices were transilluminated and viewed through a dissecting stereomicroscope. SNc and VTA were easily identifiable by location and relative transparency compared with surrounding tissue. VTA and SNc overlap along the rostrocaudal axis and appear continuous along the mediolateral axis. However, SNc is located more laterally and extends more caudally than VTA, which extends medially all the way to the midline and further rostrally than SNc (Bottjer 1993Go; Bottjer et al. 1989Go). We did not attempt to distinguish between SNc and VTA in parasagittal brain slices. In coronal slices, SNc and VTA were distinguished as shown in Fig. 1.


Figure 1
View larger version (85K):
[in this window]
[in a new window]
  		FIG. 1. A: tyrosine-hydroxylase immunostained coronal sections through the zebra finch midbrain including the substantia nigra pars compacta (SNc) and ventral tegmental area (VTA). Sections are ~250 µm apart. Dorsal is up. Scale bar is 500 µm. B: coronal hemisections corresponding approximately to the locations in (A–C) with biocytin cell fills (green; indicated by arrowheads) and tyrosine-hydroxylase (TH) immunostaining (blue). The midline is near the left edge of each image; dorsal is up. Scale bar is 500 µm. C: example of a biocytin-filled type 1 neuron (green; indicated by arrowhead) that is TH positive (blue). Scale bar is 20 µm. D: example of a biocytin-filled type 2 neuron fill that is TH negative. Scale bar is 20 µm.

 
Current-clamp recordings were obtained using the "blind" whole cell method (Blanton et al. 1989Go). Glass pipettes (5–9 M{Omega}) were filled with an internal solution containing (in mM) 120 K-methylsulfate, 10 HEPES, 0.2 EGTA, 8 NaCl, 2 MgCl, 2 ATP, and 0.3 GTP, pH ~7.3. The solution used to fill the tip of the pipette also contained 0.6% biocytin. Voltage signals were amplified 50 times and low-pass filtered at 3 kHz with a MultiClamp 700A amplifier, and data were acquired at 10 kHz with a DigiData 1322A data-acquisition system and Clampex 9.0 software (Axon Instruments, Foster City, CA). The liquid junction potential calculated with Clampex was 8.5 mV and is corrected for in figures and reported voltage values. Extracellular recordings were made in current-clamp mode using the same pipettes used for whole cell recordings except filled with 1 M NaCl; we did not attempt to "seal on" to cells. Extracellular voltage signals were amplified 1,000 times, low-pass filtered at 5 kHz, and acquired at 20 kHz.

After recording, each slice was immersion fixed in 4% paraformaldehyde in 0.1 M phosphate buffer (PB) for 12–24 h at 4°C, and subsequently cryoprotected in 30% sucrose in 0.1 M PB for ≥12 h at 4°C. Slices were then sectioned to 40 µm thickness with a freezing microtome. For TH immunostaining and biocytin cell-fill visualization, sectioned slices were incubated with 10% normal goat serum and 0.3% Triton-X in 0.1 M PB (PBT) for 1 h, and then with a rabbit anti-TH antibody (1:500; Chemicon, Temecula, CA), 1% goat serum, and 0.3% Triton-X in 0.1 M PB overnight at 4°C. Slices were subsequently washed with 0.1 M PBT three times for ≥10 min, followed by a 2-h, room-temperature incubation with Cy5- (Jackson ImmunoResearch, West Grove, PA) or Alexa 594 (Molecular Probes)-conjugated goat anti-rabbit secondary antibody (1:100) and Cy2 (Jackson)-, Alexa 488 (Molecular Probes)-, or Alexa 568 (for triple-labeling experiments following Area X tracer injections)-conjugated streptavidin (1:100) and then a second set of washes. Images were acquired using a Zeiss Axiovert 200 M microscope, Bio-Rad MRC-1024UV confocal microscope, or Olympus FV-1000 confocal microscope. Image brightness and contrast were adjusted using ImageJ (Wayne Rasband, National Institutes of Health). The TH immunostained tissue in Fig. 1A was processed using a biotinylated goat anti-rabbit secondary antibody and avidin/biotin/horseradish-peroxidase reaction using diaminobenzidine as substrate.

Data were analyzed using Clampfit 9.0 (Axon Instruments), IGOR 4.0 (Wave Metrics, Lake Oswego, OR), and Matlab (MathWorks, Natick, MA). Action potential (AP) properties (spontaneous, or in response to small depolarizing current pulses for cells not spontaneously active) are the average values for 5–40 APs. Spontaneous firing rate was calculated at the beginning of the recording as soon as it stabilized following patch rupture. Interspike-interval coefficient of variation is the SD of interspike intervals divided by their mean. Maximum "steady-state" firing rate was the highest average firing rate that could be evoked during the last 400 ms of an 800-ms step of positive current injection. AP duration was defined as the time from AP threshold (or "take off") to the return of the membrane potential to this voltage. AP amplitude was the difference between the AP peak and AP threshold voltages, and half-width is the time interval between points 50% of the AP amplitude on the rising and falling sides of the AP. Afterhyperpolarization (AHP) amplitude was the difference between AP threshold and the peak hyperpolarization after an AP. Input resistance was the slope of the current-voltage plot for voltage deflections nearest to but more negative than –60 mV. Time-dependent inward rectification ("sag") in response to hyperpolarizing current pulses and the delay to first rebound AP were quantified from traces in which the cell was hyperpolarized to a negative peak voltage of approximately –100 mV. Sag amplitude was calculated as the difference between the peak and steady-state voltage values. Delay to first rebound AP was the time from termination of the hyperpolarizing current injection to the peak of the first rebound AP. To analyze AP bursts, the onset of a burst was when the instantaneous frequency became >12 Hz and the end of the burst when instantaneous frequency fell <9 Hz [similar but not identical to Grace and Bunney (1984b)Go and Hyland et al. (2002)Go]. Cell diameters were measured in Image J and are reported as the average of the lengths of the long axis of each cell and the axis perpendicular to and centered on this axis. The width of extracellularly recorded spikes was calculated as the time from when a value 10% of the amplitude of the initial peak was reached until return to a value 10% of the amplitude of the opposite going peak that followed.

For pharmacology experiments, control firing rate was the average firing rate during the 30-s period preceding drug application. In some cells, there was an abrupt and sustained change in firing rate in the presence of a particular drug. The "drug" firing rate for these cells was calculated as the average firing rate over a 30-s period during the maximum effect of the drug because the effect of the drug often partially diminished in continued presence of the drug. In all other cells, there was no apparent effect of the drug on firing rate (a plot of instantaneous firing rate vs. time was flat). The "drug" firing rate for these cells was the average firing rate over the 30-s period 2 min after drug entered the bath. This time was chosen because drug application was approximately 2 min when no effect was apparent, and the onset of drug effect was faster than 2 min in cells in which an effect was observed. Cells in which the firing rate did not change by more than 10% in a presence of a particular drug were considered unaffected by that drug (see RESULTS).

Drugs were diluted to their final concentration in the ACSF perfusing the slice. Apamin, dopamine, picrotoxin, TTX, and quinpirole were purchased from Sigma (St. Louis, MO). DAMGO, kynurenic acid, sulpiride, ZD7288, and 4-aminopyridine (4-AP) were purchased from Tocris (St. Louis, MO).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
DISCUSSION
 GRANTS
 ACKNOWLEDGMENTS
 REFERENCES
 
Classification of cell types

We made whole cell, current-clamp recordings from 351 neurons in the SNc and VTA of adult, male zebra finches. All cells recorded in which successful biocytin cell-fills were obtained (n = 236) were located among dense TH-immunopositive neuronal cell bodies (Fig. 1). Of the neurons we recorded, 35 were in parasagittal slices where we could not clearly distinguish SNc and VTA (see METHODS), 291 were in VTA in coronal slices, and 25 were in SNc in coronal slices. Our findings were similar for VTA and SNc neurons, and these data are combined.

Two distinct cell classes were apparent based on intrinsic electrophysiological properties, pharmacological responses, and TH immunoreactivity (Table 1). In brief, "type 1" neurons (n = 180, 51% of cells recorded) had broad action potentials (APs; >2 ms), could not sustain high firing rates (>50 Hz), were strongly hyperpolarized by dopamine or the D2 dopamine-receptor agonist quinpirole but not by the µ–opioid receptor agonist DAMGO, and most (94%) were TH positive. "Type 2" neurons (n = 171, 49% of cells recorded) had relatively narrow action potentials (<2 ms), could sustain high firing rates, were hyperpolarized by DAMGO, and were TH negative. The distribution of AP durations for all cells recorded was bimodal with only a small region of overlap (Fig. 2). The other characteristic electrophysiological properties and pharmacological responses stated in the preceding text distinguished cells in the region of overlap. However, in a small number of cells, a single property (either response to dopamine, quinpirole, or DAMGO or TH immunoreactivity) contradicted several other properties that pointed toward a cell belonging to a particular cell class. These cells were classified according to the majority of their characteristic features. For instance, type 1 and type 2 neurons almost completely corresponded to dopaminergic and nondopaminergic neurons, respectively. However, seven TH negative neurons had electrophysiological and pharmacological properties like those of type 1 neurons and were therefore classified as type 1 (Fig. 2). These neurons may be nondopaminergic neurons with physiological properties similar to those of dopaminergic neurons; alternatively, the TH antibody used may fail to label a small percentage of dopaminergic neurons. TH negative neurons with physiological properties similar to those of dopaminergic neurons have also been observed in rodents (Cameron et al. 1997Go; Margolis et al. 2003Go; Ungless et al. 2004Go).


View this table:
[in this window]
[in a new window]
  		TABLE 1. Properties of zebra finch SNc and VTA neurons

 

Figure 2
View larger version (23K):
[in this window]
[in a new window]
  		FIG. 2. A: example of spontaneous firing in a type 1 neuron with a single action potential enlarged on the right side. B: example of spontaneously firing in a type 2 neuron with a single action potential (AP) enlarged on the right side. C: distribution of AP durations from all neurons recorded in 0.1-ms bins and fit to a double Gaussian. D: scatter plot of AP duration and spontaneous firing rate of all neurons recorded. Type 1 neurons are shown with filled circles and type 2 neurons with open squares. Symbols are blue for TH positive neurons, red for TH negative neurons, and black for neurons that were not successfully recovered for TH immunostaining.

 
In summary, consideration of multiple features best captured the major differences between the neurons we recorded. Nonetheless, even if we had used a single criterion such as TH immunoreactivity or AP duration to classify cells, the main conclusions of this study would be the same.

In addition to whole cell recordings, we also used extracellular recordings to monitor the spontaneous firing rate of a smaller sample of cells (n = 59) in experiments in which we wanted to avoid whole cell dialysis. We observed three general classes of extracellular spike waveforms: broad biphasic spikes with a "shoulder" during an initial, positive-going peak that was followed by one or two negative-going peaks (Fig. 3, A and B), relatively narrow biphasic spikes (C), and "monophasic" spikes with a negative going peak followed by a small positive peak that was only visible after averaging (D). Spontaneous firing of units with broad spikes (>2 ms) was inhibited by quinpirole (n = 13), whereas spontaneous firing of units with narrow spikes (<2 ms, biphasic or "monophasic") was inhibited by DAMGO (n = 17). Therefore in accordance with the classification of neurons recorded in whole cell mode described in the preceding text, we classified units with spikes >2 ms as type 1 and units with spikes <2 ms as type 2.


Figure 3
View larger version (32K):
[in this window]
[in a new window]
  		FIG. 3. A: example extracellular waveforms from 4 different neurons. The black traces are 300–1,000 consecutive APs aligned to the 1st peak, and the red trace is the average of the black traces. B: scatter plot of AP duration and spontaneous firing rate of all neurons recorded extracellularly. Type 1 neurons are shown with filled circles and type 2 neurons with open squares or triangles corresponding to biphasic and monophasic units, respectively. Symbols are blue for neurons inhibited by the D2 dopamine-receptor agonist quinpirole, red for neurons inhibited by the µ-opioid receptor agonist DAMGO, and black for neurons whose response to quinpirole or DAMGO was not tested.

 
Intrinsic membrane properties

Type 1 and 2 neurons differed in the characteristics of their APs as well as in their patterns of spontaneous and evoked activity. In addition to APs of longer duration, type 1 neurons also had higher AP thresholds and correspondingly larger AHPs than type 2 neurons on average (Table 1). Most type 1 neurons fired APs spontaneously at a slow rate with regular interspike intervals; three type 1 neurons did not fire spontaneously. Most type 2 neurons (149/171) also fired spontaneously, but at faster rates than type 1 neurons on average (Table 1; Fig. 2). Spontaneous firing was not an artifact of whole cell recording or dependent on fast synaptic transmission—spontaneous firing was present although significantly slower on average (P < 0.05, 2-tailed t-test) for both type 1 and 2 neurons when recorded extracellularly compared with whole cell [type 1: 2.3 ± 0.2 vs. 2.9 ± 0.1 (SE) Hz; type 2: 3.7 ± 0.5 vs. 6.7 ± 0.3 Hz; Fig. 3] and was not affected by simultaneous bath application of the ionotropic glutamate and GABA receptor antagonists kynurenic acid (1 mM) and picrotoxin (150 µM; type 1, n = 7; type 2, n = 7).

In response to depolarizing current injection, type 1 neurons could fire rapidly for a brief period but showed strong spike-frequency adaptation and could not sustain high firing rates (Fig. 4). The steady-state firing rate of type 1 neurons in response to depolarizing current pulses of increasing amplitude approached asymptote near 20 Hz on average (range, 10–50 Hz; Fig. 4). Stronger current injection resulted in depolarization block of AP firing. Type 2 neurons, in contrast, could sustain much higher firing rates in response to depolarizing current injection (Table 1; Fig. 4).


Figure 4
View larger version (47K):
[in this window]
[in a new window]
  		FIG. 4. A: response of a type 1 neuron (left) and a type 2 neuron (right) to a positive 150-pA current step of 800-ms duration. B: instantaneous firing rate (reciprocal of interspike intervals) during a series of 800-ms current steps in increasing increments of 50 pA from 50 to 250 pA (type 1 neuron, left) and from 50 to 400 pA (type 2 neuron, right). These are the same neurons as in A. C: steady-state (average firing rate over the last 400 ms of the current injection) and initial (or maximum; the first interval) firing rates during the current injections shown in B. D and E: mean steady-state and initial firing rates of type 1 (bullet; n = 131) and type 2 neurons ({circ}; n = 95) in response to current injections of the sizes indicated. Error bars indicate SD.

 
Hyperpolarizing current injection revealed both similarities and differences between type 1 and 2 neurons. Neurons of both types usually showed a slow, depolarizing sag in response to hyperpolarizing current pulses suggestive of the h-current (Ih; Fig. 5). This voltage sag was abolished after bath application of the specific Ih blocker ZD7288 (30 µM; type 1, n = 6; type 2, n = 8; Fig. 5, A and B). ZD7288 application caused a small decrease in the spontaneous firing rate (recorded extracellularly in the presence of kynurenic acid and picrotoxin) of all type 1 neurons tested (mean ± SD decrease of 31 ± 9%, n = 6; Fig. 5D); a similar effect of ZD7288 has been observed in mammalian dopamine neurons (Appel et al. 2003Go; Neuhoff et al. 2002Go; Okamoto et al. 2006Go; Seutin et al. 2001Go). ZD7288 had a more profound effect on type 2 neurons, decreasing spontaneous firing by 76 ± 21% (n = 8; Fig. 5D). In some of the type 2 neurons to which ZD7288 was applied (n = 3), pacemaker-like spontaneous firing not only decreased in rate but also was replaced by several-second long periods of silence flanking periods of spontaneous firing of similar duration.


Figure 5
View larger version (28K):
[in this window]
[in a new window]
  		FIG. 5. A: response of a type 1 neuron to hyperpolarizing current injection (–100 pA, 800 ms) before (black trace) and in the presence of the Ih blocker ZD7288 (gray trace). B: example of a type 2 neuron to hyperpolarizing current injection (–100 pA, 800 ms) before (black trace) and in the presence of ZD7288 (gray trace). C: cumulative distributions of sag amplitudes during hyperpolarizing current injection for type 1 (black circles) and type 2 (gray sqares) neurons. These distributions were significantly different (Kolmogorov-Smirnov test, P = 0.04). D: extracellularly recorded spontaneous firing rate of type 1 (filled black circles, n = 6) and type 2 (open gray squares, n = 8) neurons before and in the presence of ZD7288.

 
Several different types of rebound activity were observed on termination of hyperpolarizing current pulses. The firing rate of many type 1 and 2 neurons accelerated briefly after release from hyperpolarizing current injection. In some neurons of either type, rebound firing was noticeably delayed as the membrane potential slowly ramped toward AP threshold (Fig. 6A). Such delays between release from hyperpolarization and the time of the first rebound AP were more common and typically longer in type 1 neurons than in type 2 neurons (Fig. 6D). When these neurons were held at –70 to –80 mV by continuous current injection and then briefly depolarized, a ramping potential was observed and spiking was delayed (Fig. 6, B and C). This type of ramp or delay response can be caused by a strong A-type potassium current, which can be blocked nonselectively by 4-AP. After bath application of a relatively low concentration of 4-AP (100 µM), type 2 neurons with a voltage ramp and delay to fire when depolarized from –80 mV fired near the beginning of depolarizing current steps with no noticeable voltage ramp (n = 9; Fig. 6C). In contrast, in type 1 neurons, the voltage ramp and delay to fire were not affected by 100 µM (n = 6) or 1 mM (n = 3) 4-AP; they were, however, suppressed by 4 mM 4-AP (n = 3; Fig. 6B). As expected, 4-AP seemed to block several potassium conductances and caused the appearance of large, spontaneous postsynaptic potentials; hence, experiments with millimolar concentrations of 4-AP were performed in the presence of kynurenic acid and picrotoxin. Our observations are consistent with voltage-clamp experiments in rat VTA neurons showing stronger A currents with lower 4-AP sensitivity in dopaminergic neurons compared with nondopaminergic neurons (Koyama and Appel 2006Go; see also Liss et al. 2001Go).


Figure 6
View larger version (25K):
[in this window]
[in a new window]
  		FIG. 6. A: example of a type 1 neuron that showed a long delay to fire action potentials after termination of hyperpolarizing current injection (–75 pA, 800 ms). B: response of the same type 1 neuron shown in A, held near –70 mV by continuous current injection, to depolarizing current injection (75 pA, 1000 ms) before (black trace) and in the presence of the A-type potassium current blocker 4-aminopyridine (4-AP; 4 mM). C: response of a type 2 neuron, held near –80 mV by continuous current injection, to depolarizing current injection (100 pA, 800 ms) before (black trace) and in the presence of 0.1 mM 4-AP. D: cumulative distributions of time delays from termination of hyperpolarizing current injection to first action potential in type 1 (black circles) and type 2 (gray squares) neurons. These distributions were significantly different (Kolmogorov-Smirnov test, P < 0.001).

 
Type 1 neurons often rebounded from hyperpolarizing current injection with large, repetitive membrane potential oscillations (Fig. 7A). These "slow oscillatory potentials" (SOPs) were also commonly observed when depolarizing cells from a holding potential near –80 mV (Fig. 7, C and D), and overall they were observed in 60% of type 1 neurons. SOPs occurred at a rate similar to the spontaneous AP firing rate and persisted in the presence of the sodium channel blocker TTX (2 µM, n = 13; Fig. 7, C and D). SOPs were abolished when extracellular Ca2+ was replaced with equimolar Mg2+ (n = 10; Fig. 7C). Furthermore, apamin (100 nM), which blocks SK calcium-dependent potassium channels, prolonged the repolarization phase of the oscillation, converting SOPs to longer duration oscillations or plateau potentials (n = 7; Fig. 7D).


Figure 7
View larger version (32K):
[in this window]
[in a new window]
  		FIG. 7. A: example of a type 1 neuron that showed repetitive membrane potential oscillations after injection of –25 pA for 800 ms. B: example of a type 1 neuron that fired short bursts of APs repetitively after injection of –125 pA for 800 ms. C, left: example of a type 1 neuron, held near –80 mV by continuous current injection, in response to depolarizing current injection (30 pA, 800 ms) in the presence of TTX (black trace) or TTX and no extracellular Ca2+ (gray trace). Right: amplitude of the maximum membrane potential fluctuation in type 1 neurons recorded in the presence of TTX (control) or TTX and no extracellular Ca2+ (n = 10 cells). To calculate this amplitude for control conditions (TTX), we used traces in which the cell was held near –80 mV and a current step of minimum magnitude to evoke slow oscillatory potentials (SOPs) was applied; an equivalent current step was used for the no extracellular Ca2+ condition. For control, the amplitude was the difference in membrane potential between the peak of the 1st oscillation and the following trough. For 0 extracellular Ca2+ (where there were no clear SOPs), the amplitude was the difference between the peak membrane potential during the current injection and the minimum value that occurred after that peak. D, left: example of a type 1 neuron, held near –80 mV by continuous current injection, in response to depolarizing current injection (60 pA, 800 ms) in the presence of TTX (black trace) or TTX and apamin (gray trace). Right: half-width of SOPs in type 1 neurons recorded in the presence of TTX (control) or TTX and apamin (n = 7 cells); the y axis is on a log scale. To calculate half-width, the cell was held near –80 mV and a current step of minimum magnitude to evoke SOPs was applied. Baseline was set as the trough after the peak of the first SOP for control traces (TTX) or, for apamin traces, as the membrane potential just before the end of the current injection (see example trace). Half-width of the SOP preceding baseline is then the time interval between points 50% of the SOP peak (relative to baseline) on the rising and falling sides of the SOP. E: example of a type 2 neuron that fired a single burst of action potentials on rebound from hyperpolarizing current injection (–50 pA, 800 ms). F: example of a Type 2 neuron that fired a burst of action potentials riding on a long, depolarized plateau potential after hyperpolarizing current injection (–100 pA, 800 ms). G, left: example of a type 2 neuron, held near –80 mV by continuous current injection, in response to depolarizing current injection (40 pA, 800 ms) in control conditions (black trace) or when extracellular Ca2+ was removed (gray trace). Right: amplitude of the maximum membrane potential fluctuation (see METHODS) in type 2 neurons recorded in control conditions or when extracellular Ca2+ was removed (n = 7 neurons). Amplitude was calculated the same way as for SOPs described in the preceding text, except traces were smoothed using a Boxcar filter to remove any action potentials.

 
SOPs were subthreshold or had APs rising before or at the peak of the underlying membrane potential oscillation. In some type 1 neurons (n = 19, 11%), however, bursts of APs rode on top of SOPs (Fig. 7B). These bursts consisted of two to five spikes except for one neuron that exhibited longer oscillations carrying bursts of 2–11 APs. Spontaneous bursts (not associated with current injection) were also observed in one type 1 cell. Across cells (n = 19), the mean frequency of APs in bursts was 40 ± 16 Hz (range: 15–71 Hz).

Subthreshold membrane potential oscillations were never observed in type 2 neurons, including six type 2 neurons bathed in TTX. A subset of type 2 neurons (n = 37, 22%) did, however, exhibit single bursts of APs after hyperpolarizing current pulses or when depolarized from a holding potential near –80 mV (Fig. 7, E and G). These bursts were similar to those caused by low-threshold Ca2+ spikes (LTS) in several classes of mammalian neurons. Consistent with this comparison, the burst of APs and depolarizing "bump" underlying these bursts in type 2 neurons were abolished when extracellular Ca2+ was replaced with equimolar Mg2+ (n = 7). Ten type 2 neurons (6%) exhibited rebound bursts riding on top of depolarized potentials of particularly long duration (Fig. 7F). Five of these neurons stopped firing spontaneous APs or alternated spontaneous firing with periods (several seconds long) of rest at a relatively depolarized potential (–40 mV). All of the 10 neurons with long LTSs except for one did not fire repetitively in response to depolarizing current injection; instead they fired just one or a few APs at the beginning of the current pulse.

Pharmacological responses

We tested the effect of dopamine, the D2-dopamine receptor agonist quinpirole, and the µ-opioid receptor agonist DAMGO on zebra finch SNc and VTA neurons because these agents have been useful to distinguish different types of neurons in the rodent SNc and VTA (see DISCUSSION).

Bath application of 50 µM dopamine or 10 µM quinpirole hyperpolarized and strongly inhibited spontaneous firing of all type 1 neurons tested (n = 28 and 62, respectively; Fig. 8). The decrease in spontaneous firing rate in the presence of dopamine or quinpirole was >50% in all type 1 neurons tested, and most of these cells completely stopped firing spontaneous APs (n = 23/28 and 49/62). The effects of dopamine and quinpirole were reversible, although wash-out of quinpirole was slower. The effect of dopamine or quinpirole on spontaneous firing of type 1 neurons could be reversed by subsequent bath application of the D2 dopamine receptor antagonist sulpiride (10 µM, n = 6 and 3, respectively; Fig. 8A), further implicating D2-like receptors in mediating the effect of dopamine and quinpirole. Inhibition of spontaneous firing and hyperpolarization were likely caused by increased potassium conductance, as the intersection of current-voltage plots obtained before and in the presence of dopamine were near the estimated potassium equilibrium potential (n = 2, –103 and –92 mV; Nernst EK+ = –100 mV).


Figure 8
View larger version (54K):
[in this window]
[in a new window]
  		FIG. 8. A: example of a type 1 neuron that was hyperpolarized and stopped firing in response to bath application of dopamine (DA). The effect of dopamine could be reversed by subsequent application of the D2 dopamine receptor agonist sulpiride. Top portion: membrane potential; bottom portion: instantaneous firing rate over the time course of the experiment. B: example of a type 1 neuron that did not change its firing rate in response to bath application of the µ-opioid receptor agonist DAMGO but was hyperpolarized and stopped firing in response to the D2 dopamine receptor agonist quinpirole. C: example of a type 2 neuron that was hyperpolarized and stopped firing in response to DAMGO and the firing rate of which was not affected by quinpirole. D–F: scatter plots of the firing rates of type 1 ({square}) and type 2 ({triangleup}) neurons before (control) and in the presence of dopamine (D), quinpirole (E), or DAMGO (F). - - -, unity (no change in firing rate).

 
In contrast to its strong effect on type 1 neurons, dopamine had no effect on the spontaneous firing rate of the majority of type 2 neurons tested (n = 14/24; Fig. 8D). However, dopamine reversibly increased (n = 3/24) or decreased (n = 7/24) the firing rate of some type 2 neurons. The firing rate of four type 2 neurons was moderately reduced (16–44%) by dopamine and three other type 2 neurons stopped firing spontaneous APs in the presence of dopamine; all other properties of these neurons that we tested were characteristic of type 2 neurons (narrow APs, ability to sustain fast firing, inhibition by DAMGO, and lack of TH staining). Like dopamine, quinpirole had no effect on the spontaneous firing rate of most (n = 26/29) type 2 neurons (Fig. 8E). Spontaneous firing of three type 2 neurons was weakly inhibited (11–30%) by quinpirole.

Bath application of 3 µM DAMGO had no effect on the spontaneous firing rate of most type 1 neurons (n = 42/43; Fig. 8F); spontaneous firing of one type 1 neuron was reversibly inhibited by 28% in the presence of DAMGO. In contrast, DAMGO reversibly hyperpolarized and reduced the spontaneous firing rate of all type 2 neurons tested (n = 53; Fig. 8, C and F). Spontaneous firing was completely suppressed by DAMGO in most type 2 neurons (n = 40/53).

The effects of both quinpirole and DAMGO were tested in 40 type 1 neurons and 29 type 2 neurons. Similar effects were observed regardless of the order in which drugs were applied. In these experiments, all type 1 neurons but one (39/40) were inhibited by quinpirole but not DAMGO, and all but three type 2 neurons (26/29) were inhibited by DAMGO but not quinpirole. Thus we observed only 4 neurons that were inhibited by both quinpirole and DAMGO, whereas 65 neurons were inhibited by only one or the other drug.

Inhibition (or lack thereof) by dopamine, quinpirole, and DAMGO were good predictors of TH immunostaining: 19/22 neurons that were inhibited by dopamine were TH positive and 16/16 neurons that were not inhibited by dopamine were TH negative; 39/43 neurons that were inhibited by quinpirole were TH positive, and 16/16 neurons that were not inhibited by quinpirole were TH negative; 37/37 neurons that were inhibited by DAMGO were TH negative, and 29/29 neurons that were not inhibited by DAMGO were TH positive.

We also found that bath application of the GABA-B receptor agonist baclofen (50 µM) completely suppressed spontaneous firing and hyperpolarized all type 1 (n = 5) and type 2 (n = 5) neurons tested (data not shown).

Recordings from identified Area X-projecting neurons and from juvenile zebra finches

It is possible that few of our recordings were from neurons that project to Area X of the song control system and that the physiological properties of Area X-projecting neurons differ in some way from those of the VTA neurons we did record from. Therefore in some experiments we recorded from slices taken from birds in which we had made bilateral injections of retrograde tracer (fluorescently labeled cholera toxin subunit B) into Area X 5–9 days before slicing. We recorded from 101 neurons from 20 birds that had bilateral injections of tracer within the borders of Area X. Of these neurons, 12 of 53 type 1 neurons (23%) were retrogradely labeled (and were also TH positive; Fig. 9), whereas zero of 48 type 2 neurons were retrogradely labeled. The 12 identified Area-X projecting neurons did not qualitatively or quantitatively differ from our larger data set of type 1 neurons in any of the physiological properties or pharmacological responses that we measured (Table 1). In fact, this experiment suggests that ~20% or more of all the type 1 neurons we recorded from project to Area X (and probably a greater percentage given that our injections did not completely fill Area X). Moreover, this experiment suggests that much fewer (if any) type 2 neurons than type 1 neurons project to Area X.


Figure 9
View larger version (103K):
[in this window]
[in a new window]
  		FIG. 9. Example of a biocytin-filled type 1 neuron (A) that was also retrogradely labeled from a cholera toxin injection in Area X (B) and TH immunopositive (C). D: overlay of the images in A–C. Inset: cholera toxin injection site (saturated pixels) and the borders of Area X.

 
It is also possible that the physiological properties of SNc and VTA neurons differ in juvenile birds during the developmental period in which song learning occurs. Hence, we recorded from slices from three male zebra finches that were 30–35 days post hatch (when they are sensitive to memorizing the song of adult birds) and two male zebra finches that were 47–52 days post hatch (during which vocal learning has already begun). We recorded from 12 type 1 neurons and 14 type 2 neurons from the birds 30–35 days post hatch, and 7 type 1 neurons and 4 type 2 neurons from the birds 47–52 days post hatch. The physiological properties and pharmacological responses that we measured from these neurons did not qualitatively or quantitatively differ from those of adult type 1 and type 2 neurons (Table 1). One exception is that one juvenile type 1 neuron was strongly hyperpolarized by both quinpirole and DAMGO.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
GRANTS
 ACKNOWLEDGMENTS
 REFERENCES
 
The zebra finch SNc and VTA neurons described here are in many ways very similar in their electrophysiological and pharmacological properties to SNc and VTA neurons described in mammals (Cameron et al. 1997Go; Grace and Onn 1989Go; Johnson and North 1992Go; Lacey et al. 1989Go; Mueller and Brodie 1989Go; Richards et al. 1997Go; Yung et al. 1991Go). In rodent brain slices, three types of SNc and VTA neurons—principal, secondary, and tertiary—are distinguished by electrophysiological properties and pharmacological responses. Type 1 neurons in the zebra finch SNc and VTA are very similar to principal neurons, which comprise ~60% of rodent VTA neurons. Both type 1 (zebra finch) and principal (rodent) neurons have long duration APs (>2 ms), have relatively high AP thresholds and large AHPs, show slow spontaneous firing (mean: <4 Hz, range: 1–8 Hz) that appears to generated by calcium- and apamin-sensitive membrane-potential oscillations (SOPs), show strong spike-frequency adaptation in response to depolarizing input, are inhibited by dopamine via D2 dopamine receptors, are not inhibited by DAMGO, and are dopaminergic (>90% are TH positive). Additionally, these neurons commonly exhibit a depolarizing sag in response to hyperpolarization attributable to Ih and, in some neurons, prominently delayed rebound firing caused by an A-type potassium current. Although the burst firing we recorded in some type 1 neurons is not normally observed in rodent principal neurons in vitro, burst firing is a prominent mode of firing in these neuron in vivo and is observed in vitro under certain conditions, such as when nickel- and mibefradil-sensitive voltage-gated calcium channels or apamin-sensitive calcium-activated potassium channels are blocked (Shepard and Bunney 1991Go; Wolfart and Roeper 2002Go). Thus these comparisons suggest the main physiological features of mammalian and zebra finch dopaminergic neurons appear to have been conserved during vertebrate evolution.

However, principal neurons are not the only dopaminergic neurons in the rodent SNc and VTA. Tertiary neurons make up ~30% of rodent VTA neurons, and ~30% of tertiary neurons are TH positive. Tertiary neurons share all of the properties of principal neurons and type 1 neurons described in the preceding text with one exception: they are hyperpolarized by DAMGO. We did not find strong evidence for tertiary-like neurons in the zebra finch SNc or VTA. Only one of the neurons we recorded that had electrophysiological properties like those of principal and tertiary neurons (type 1 neurons) was inhibited (relatively weakly) by DAMGO. Further, the neurons we recorded that were strongly inhibited by DAMGO (type 2 neurons), were all TH negative (unlike tertiary neurons as a population), and had electrophysiological properties that differed from those of principal and tertiary neurons.

Indeed, type 2 neurons had electrophysiological properties that correspond more closely to those of secondary neurons of the rodent SNc and VTA. Both type 2 neurons in zebra finches and secondary neurons in rodents have relatively narrow APs, are silent or exhibit relatively fast spontaneous firing, can sustain high-frequency firing in response to depolarizing input, sometimes display LTSs, are inhibited by DAMGO, and are nondopaminergic. Also like secondary neurons, most type 2 neurons were not hyperpolarized by dopamine or quinpirole. Differing from secondary neurons, however, type 2 neurons exhibited a slowly developing depolarizing sag attributable to Ih in response to hyperpolarizing current injection. Apparently this difference is not minor, as spontaneous firing of type 2 neurons was considerably disrupted by the Ih blocker ZD7288. Another difference between type 2 and secondary neurons is that whereas secondary neurons comprise only 10–15% of rodent SNc and VTA neurons, type 2 neurons were encountered nearly as frequently as type 1 neurons (49% of cells recorded). This may reflect differences in technique (whole cell vs. sharp electrode recording, visualized vs. blind recording) or slice conditions, or it may well indicate that type 2 neurons are proportionally more numerous in zebra finches than secondary neurons in rodents.

In summary, the zebra finch SNc and VTA appear to contain two types of neurons—type 1 and type 2—corresponding to principal and secondary neurons, but not tertiary neurons, of the rodent SNc and VTA. The anatomical projections and functions of tertiary neurons and the neurotransmitter(s) used by nondopaminergic tertiary neurons are presently unknown. The significance of the apparent absence of this class of neuron in the zebra finch SNc and VTA is therefore uncertain.

In contrast to type 1 neurons, we do not know what neurotransmitter(s) type 2 neurons use or where they project anatomically. The comparable class of cells in rodents, secondary neurons, are GABAergic and are thought to make inhibitory synapses on dopaminergic neurons. In rodent brain slices, DAMGO reduces the frequency but not amplitude of TTX-sensitive spontaneous inhibitory postsynaptic potentials (IPSPs) in dopaminergic neurons, presumably by inhibiting AP firing in secondary neurons (Johnson and North 1993Go). Hence, via disinhibition, DAMGO can increase the firing rate of dopaminergic neurons (Johnson and North 1993Go; Margolis et al. 2003Go). We did not observe an increase in the spontaneous firing rate of any type 1 neurons in response to DAMGO (Fig. 8). Moreover, we rarely observed spontaneous IPSPs [or inhibitory postsynaptic currents (IPSCs) in voltage clamp] in type 1 or 2 neurons. Usually all spontaneous PSCs were miniature excitatory postsynaptic currents (insensitive to TTX and blocked by the AMPA receptor antagonist CNQX; data not shown). These results may be explained if axons were severed during slicing or if putative type 2 neurons making intra-VTA synapses were silent (in 1 study, increased extracellular potassium was necessary to evoke spontaneous IPSPs in rodent dopaminergic neurons) (Johnson and North 1993Go).

At least some GABAergic (presumably secondary) neurons in the rodent VTA project to striatum or to prefrontal cortex (Carr and Sesack 2000Go; Swanson 1982Go; Van Bockstaele and Pickel 1995Go). The functions of these projections are unknown. TH-negative neurons in the SNc and VTA that project to nuclei of the song system motor pathway have been reported in retrograde tracing studies (Appeltants et al. 2000Go, 2002Go); these could be type 2 neurons and/or TH-negative type 1 neurons. Nondopaminergic neurons might also contribute to the substantial projection from VTA to Area X. It is additionally worth noting that there are neurons in the mammalian SNc and VTA that use glutamate as a neurotransmitter as well, and at least some of these glutamatergic neurons might also be dopaminergic (Chuhma et al. 2004Go; Lavin et al. 2005Go; Sulzer et al. 1998Go).

The similarities between zebra finch and mammalian dopaminergic neurons are of practical importance. First, the criteria used to identify mammalian dopaminergic neurons during extracellular recordings in vivo—relatively slow spontaneous firing rate, long-duration action potential waveforms, and inhibition by D2 dopamine-receptor agonists—seem applicable in zebra finches as well (but see Hyland et al. 2002Go; Kiyatkin and Rebec 1998Go; Ungless et al. 2004Go). Second, intrinsic membrane properties strongly influence the output of neurons. It is likely that zebra finch and mammalian dopaminergic neurons show similar patterns and rates of activity in vivo. Specifically, mammalian dopaminergic neurons exhibit slow spontaneous activity (<10 Hz) and fire short bursts of APs "spontaneously" and in response to unexpected rewards, reward-predicting stimuli, or other surprising or novel stimuli (Horvitz 2000Go; Hyland et al. 2002Go; Schultz 1998Go). In zebra finch dopaminergic neurons in vitro, the spontaneous firing rate and the firing rate during depolarizing current injection or intrinsic bursts were markedly similar to rates observed during tonic and burst firing of dopaminergic neurons in anesthetized or awake rats and monkeys (Grace and Bunney 1984aGo,bGo; Hyland et al. 2002Go; Schultz 1998Go). This is probably due to shared intrinsic mechanisms that drive spontaneous activity (SOPs) and strongly constrain responses to excitatory input.

Our finding of comparable physiological and pharmacological properties of avian and mammalian SNc/VTA neurons is perhaps not surprising given the physiological similarities observed between several classes of avian and mammalian neurons in basal ganglia structures (Farries and Perkel 2000Go, 2002Go) and the similar anatomical organization of avian and mammalian dopaminergic systems. Yet these comparisons are still somewhat striking given that the last common ancestor of mammals and birds lived >300 million years ago. The specific functions of dopaminergic and other SNc and VTA neurons remain uncertain. Based on similarities in the anatomical projections and physiological properties of avian and mammalian dopaminergic neurons, and similar physiological and behavioral effects of dopaminergic agents, it is reasonable to hypothesize that at least some functions of dopamine are conserved across vertebrates. Because song learning and production are controlled by well-defined neural circuits devoted to a single behavior, research in songbirds may offer unique insights into how dopamine in basal ganglia and other structures contributes to learning and performing complex motor behaviors (Doupe et al. 2005Go).


   GRANTS
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 GRANTS
ACKNOWLEDGMENTS
 REFERENCES
 
This work was supported by National Institutes of Health Grants MH-066128, T32-DC-05361, and P30 DC-04661 and a National Science Foundation Graduate Fellowship.


   ACKNOWLEDGMENTS
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 GRANTS
 ACKNOWLEDGMENTS
REFERENCES
 
We thank Drs. M. Solis and A. Weaver for helpful comments on the manuscript and G. MacDonald for advice on microscopy.


   FOOTNOTES
 
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Address for reprint requests and other correspondence: S. D. Gale, University of Washington, Dept. of Otolaryngology, Box 356515, Seattle, WA 98195 (E-mail: samgale{at}u.washington.edu)


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 GRANTS
 ACKNOWLEDGMENTS
 REFERENCES
 
Appel SB, Liu Z, McElvain MA, and Brodie MS. Ethanol excitation of dopaminergic ventral tegmental area neurons is blocked by quinidine. J Pharmacol Exp Ther 306: 437–446, 2003.[Abstract/Free Full Text]

Appeltants D, Absil P, Balthazart J, and Ball GF. Identification of the origin of catecholaminergic inputs to HVc in canaries by retrograde tract tracing combined with tyrosine hydroxylase immunocytochemistry. J Chem Neuroanat 18: 117–133, 2000.[CrossRef][ISI][Medline]

Appeltants D, Ball GF, and Balthazart J. The origin of catecholaminergic inputs to the song control nucleus RA in canaries. Neuroreport 16: 649–653, 2002.

Blanton MG, Lo Turco JJ, and Kriegstein AR. Whole cell recording from neurons in slices of reptilian and mammalian cerebral cortex. J Neurosci Methods 30: 203–210, 1989.[CrossRef][ISI][Medline]

Bottjer SW. The distribution of tyrosine hydroxylase immunoreactivity in the brains of male and female zebra finches. J Neurobiol 24: 51–69, 1993.[CrossRef][ISI][Medline]

Bottjer SW, Halsema KA, Brown SA, and Miesner EA. Axonal connections of a forebrain nucleus involved with vocal learning in zebra finches. J Comp Neurol 279: 312–326, 1989.[CrossRef][ISI][Medline]

Bottjer SW, Miesner EA, and Arnold AP. Forebrain lesions disrupt development but not maintenance of song in passerine birds. Science 224: 901–903, 1984.[Abstract/Free Full Text]

Brainard MS and Doupe AJ. Interruption of a basal ganglia-forebrain circuit prevents plasticity of learned vocalizations. Nature 404: 762–766, 2000.[CrossRef][Medline]

Cameron DL, Wessendorf MW, and Williams JT. A subset of ventral tegmental area neurons is inhibited by dopamine, 5-hydroxytryptamine and opioids. Neuroscience 77: 155–166, 1997.[CrossRef][ISI][Medline]

Carr DB and Sesack SR. GABA-containing neurons in the rat ventral tegmental area project to the prefrontal cortex. Synapse 38: 114–123, 2000.[CrossRef][ISI][Medline]

Chuhma N, Zhang H, Masson J, Zhuang X, Sulzer D, Hen R, and Rayport S. Dopamine neurons mediate a fast excitatory signal via their glutamatergic synapses. J Neurosci 24: 972–981, 2004.[Abstract/Free Full Text]

Ding L and Perkel DJ. Dopamine modulates excitability of spiny neurons in the avian basal ganglia. J Neurosci 22: 5210–5218, 2002.[Abstract/Free Full Text]

Ding L and Perkel DJ. Long-term potentiation in an avian basal ganglia nucleus essential for vocal learning. J Neurosci 24: 488–494, 2004.[Abstract/Free Full Text]

Ding L, Perkel DJ, and Farries MA. Presynaptic depression of glutamatergic synaptic transmission by D1-like dopamine receptor activation in the avian basal ganglia. J Neurosci 23: 6086–6095, 2003.[Abstract/Free Full Text]

Doupe AJ, Perkel DJ, Reiner A, and Stern EA. Birdbrains could teach basal ganglia research a new song. Trends Neurosci 28: 353–363, 2005.[CrossRef][ISI][Medline]

Durstewitz D, Kroner S, and Güntürkün O. The dopaminergic innervation of the avian telencephalon. Prog Neurobiol 59: 161–195, 1999.[CrossRef][ISI][Medline]

Farries MA and Perkel DJ. Electrophysiological properties of avian basal ganglia neurons recorded in vitro. J Neurophysiol 84: 2502–2513, 2000.[Abstract/Free Full Text]

Farries MA and Perkel DJ. A telencephalic nucleus essential for song learning contains neurons with physiological characteristics of both striatum and globus pallidus. J Neurosci 22: 3776–3787, 2002.[Abstract/Free Full Text]

Gale SD and Perkel DJ. Properties of dopamine release and uptake in the songbird basal ganglia. J Neurophysiol 93: 1871–1879, 2005.[Abstract/Free Full Text]

Grace AA and Bunney BS. The control of firing pattern in nigral dopamine neurons: single spike firing. J Neurosci 4: 2866–2876, 1984a.[Abstract]

Grace AA and Bunney BS. The control of firing pattern in nigral dopamine neurons: burst firing. J Neurosci 4: 2877–2890, 1984b.[Abstract]

Grace AA and Onn, S-P. Morphology and electrophysiological properties of immunocytochemically identified rat dopamine neurons recorded in vitro. J Neurosci 9: 3463–3481, 1989.[Abstract]

Harding CF, Barclay SR, and Waterman SA. Changes in catecholamine levels and turnover rates in hypothalamic, vocal control, and auditory nuclei in male zebra finches during development. J Neurobiol 34: 329–346, 1998.[CrossRef][ISI][Medline]

Horvitz JC. Mesolimbocortical and nigrostriatal dopamine responses to salient non-reward events. Neuroscience 96: 651–656, 2000.[CrossRef][ISI][Medline]

Hyland BI, Reynolds JNJ, Hay J, Perk CG, and Miller R. Firing modes of midbrain dopamine cells in the freely moving rat. Neuroscience 114: 475–492, 2002.[CrossRef][ISI][Medline]

Johnson SW and North RA. Two types of neurone in the rat ventral tegmental area and their synaptic inputs. J Physiol 450: 455–468, 1992.[Abstract/Free Full Text]

Johnson SW and North RA. Opioids excite dopamine neurons by hyperpolarization of local interneurons. J Neurosci 12: 483–488, 1993.

Kiyatkin EA and Rebec GV. Heterogeneity of ventral tegmental area neurons: single-unit recordings and iontophoresis in awake, unrestrained rats. Neuroscience 85: 1285–1309, 1998.[CrossRef][ISI][Medline]

Koyama S and Appel SB. A-type K+ current of dopamine and GABA neurons in the ventral tegmental area. J Neurophysiol in press (Epub ahead of print), 2006.

Lacey MG, Mercuri NB, and North RA. Two cell types in rat substantia nigra zona compacta distinguished by membrane properties and the actions of dopamine and opioids. J Neurosci 9: 1233–1241, 1989.[Abstract]

Lavin A, Nogueira L, Lapish CC, Wightman RM, Phillips PE, and Seamans JK. Mesocortical dopamine neurons operate in distinct temporal domains using multimodal signaling. J Neurosci 25: 5013–5023, 2005.[Abstract/Free Full Text]

Lewis JW, Ryan SM, Arnold AP, and Butcher LL. Evidence for a catecholaminergic projection to Area X in the zebra finch. J Comp Neurol 196: 347–354, 1981.[CrossRef][ISI][Medline]

Liss B, Franz O, Sewing S, Bruns R, Neuhoff H, and Roeper J. Tuning pacemaker frequency of individual dopaminergic neurons by Kv4.3L and KChip3.1 transcription. EMBO J 20: 5715–5724, 2001.[CrossRef][ISI][Medline]

Margolis EB, Hjelmstad GO, Bonci A, and Fields HL. Kappa-opioid agonists directly inhibit midbrain dopaminergic neurons. J Neurosci 23: 9981–9986, 2003.[Abstract/Free Full Text]

Mueller AL and Brodie MS. Intracellular recording from putative dopamine-containing neurons in the ventral tegmental area of Tsai in a brain slice preparation. J Neurosci Methods 28: 15–22, 1989.[CrossRef][ISI][Medline]

Neuhoff H, Neu A, Liss B, and Roeper J. I(h) channels contribute to the different functional properties of identified dopaminergic subpopulations in the midbrain. J Neurosci 22: 1290–1302, 2002.[Abstract/Free Full Text]

Okamoto T, Harnett MT, and Morikawa H. Hyperpolarization-activated cation current (Ih) is an ethanol target in midbrain dopamine neurons of mice. J Neurophysiol 95: 619–626, 2006.[Abstract/Free Full Text]

Reiner A, Karle EJ, Anderson KD, and Medina L. Catecholaminergic perikarya and fibers in the avian nervous system. In: Phylogeny and Development of Catecholaminergic Systems in the CNS of Vertebrates, edited by Smeets WJAJ and Reiner A. Cambridge, UK: Cambridge Univ. Press, 1994. p. 135–181.

Reiner A, Perkel DJ, Bruce LL, Butler AB, Csillag A, Kuenzel W, Medina L, Paxinos G, Shimizu T, Striedter G, Wild M, Ball GF, Durand S, Güntürkün O, Lee DW, Mello CV, Powers A, White SA, Hough G, Kubikova L, Smulders TV, Wada K, Dugas-Ford J, Husband S, Yamamoto K, Yu J, Siang C, and Jarvis ED. Avian Brain Nomenclature Forum. Revised nomenclature for avian telenchephalon and some related brain stem nuclei. J Comp Neurol 473: 377–414, 2004.[CrossRef][ISI][Medline]

Richards CD, Shiroyama T, and Kitai ST. Electrophysiological and immunocytochemical characterization of GABA and dopamine neurons in the substantia nigra of the rat. Neuroscience 80: 545–557, 1997.[CrossRef][ISI][Medline]

Scharff C and Nottebohm F. A comparative study of the behavioral deficits following lesions of various parts of the zebra finch song system: implications for vocal learning. J Neurosci 11: 2896–2913, 1991.[Abstract]

Schultz W. Predictive reward signal of dopamine neurons. J Neurophysiol 80: 1–27, 1998.[Abstract/Free Full Text]

Schultz W. Getting formal with dopamine and reward. Neuron 36: 241–263, 2002.[CrossRef][ISI][Medline]

Seutin V, Massotte L, Renette MF, and Dresse A. Evidence for a modulatory role of Ih on the firing of a subgroup of midbrain dopamine neurons. Neuroreport 12: 255–258, 2001.[CrossRef][ISI][Medline]

Shepard PD and Bunney BS. Repetitive firing properties of putative dopamine-containing neurons in vitro: regulation by an apamin-sensitive Ca2+-activated K+ conductance. Exp Brain Res 86: 141–150, 1991.[ISI][Medline]

Soha JA, Shimizu T, and Doupe AJ. Development of the catecholaminergic innervation of the song system of the male zebra finch. J Neurobiol 29: 473–489, 1996.