Animals used in this study were treated in strict accordance with the American Physiological Society’s Guiding Principles in the Care and Use of Animals and the US National Institutes for Health Guide for the Care and Use of Laboratory Animals; the protocol for all experimental methods was approved by the Institutional Animal Care Committee of the University of Illinois at Chicago. Both male and female Fisher 344 rats (14 - 18 days old) were decapitated and the brain quickly removed. The brain was placed in an ice-cold cutting solution (in mM: 220 sucrose, 2.5 KCl, 2.4 CaCl₂, 1.3 MgSO₄, 1.24 NaH₂PO₄, 26 NaHCO₃, 11 _D-glucose), which was constantly bubbled with 95% O₂ and 5% CO₂. Transverse brain slices (350 - 400 μm thick) were made on a Vibratome (Series 1000 plus, St. Louis, MO, USA). The brain slices were incubated for 3 - 4 hr in an artificial cerebrospinal fluid (ACSF) (in mM: 126 NaCl, 2.5 KCl, 2.4 CaCl₂, 1.3 MgSO₄, 1.24 NaH₂PO₄, 26 NaHCO₃, 11 _D-glucose, osmolarity 300 mOsm), which was constantly bubbled with 95% O₂ and 5% CO₂ at 26˚C. VTA neurons were dissociated as previously described [22] . Specifically, brain slices were next incubated in a HEPES-buffered solution (see below) containing papain (15 - 18 U/ml) at 32˚C for 20 - 25 min. After papain treatment, the brain slices were further incubated in the ACSF for 20 - 40 min. The VTA neurons were dissociated by a vibrating stylus apparatus for dispersing cells from the brain slices. First the brain slice was transferred to a poly-_D-lysine- coated 35 mm culture dish (Becton Dickenson, Bedford, MA, USA) containing the HEPES-buffered solution. A grid of nylon threads glued to a U-shaped metal frame was used to hold the brain slice down during cell dissociation. After the VTA was visually identified, the vibrating stylus was placed in the appropriate region with a micromanipulator. The stylus was made of glass capillary tubing (1.5 mm o.d.) pulled to a fine tip, fire-polished (200 - 400 μm in diameter) and mounted on the vibrating apparatus, which horizontally vibrated the stylus tip (excursions of 100 - 200 μm at 20 - 25 Hz). Once the cell dissociation procedure was completed (4 - 7 min), the brain slice was removed from the culture dish, and the dissociated neurons settled and adhered to the bottom of the dish within 20 min.

Electrophysiological measurements were made with an Axopatch-1B patch-clamp amplifier (Axon Instruments, Union City, CA, USA). Microelectrodes were fabricated on a P-97 puller (Sutter Instrument Company, Novato, CA, USA), from LE16 glass capillaries (Dagan, Minneapolis, MN, USA) and heat-polished on a microforge (Narishige, Tokyo, Japan). The tip resistances of the electrodes were 3 - 7 MΩ when filled with a pipette solution (in mM: 60 K-acetate, 60 KCl, 1 CaCl₂, 2 MgCl₂, 40 HEPES; pH 7.2 adjusted with KOH, final [K⁺]_i = 131 mM; osmolality 290 mOsm). Nystatin-perforated patch recording was used to minimize the dialysis of intracellular contents and therefore prevent the rundown of I_M, as previously described [22] . Nystatin was dissolved in me- thanol at a concentration of 10 mg/ml. This nystatin stock solution was diluted with the pipette solution to a final concentration of 100 - 200 μg/ml and the electrodes were backfilled with this solution. After the cell-attached configuration was attained, the access resistance was periodically monitored and capacitive transients were cancelled. When the access resistance had reached a steady level (15 - 30 MΩ), the recording was started. In case of the sudden change of the access resistance, the recording was stopped. Voltage-clamp recording was done in a HEPES-buffered solution (in mM: 145 NaCl, 2.5 KCl, 2 CaCl₂, 1 MgCl₂, 10 HEPES and 11 _D-glucose; pH adjusted to 7.4 with NaOH; osmolarity 300 mOsm) constantly bubbled with 100% O₂. The liquid junction potential between the pipette solution and the HEPES-buffered solution was estimated to be 5 mV [23] and the results have been corrected by this amount. Membrane currents and voltage were filtered at 1 kHz by a −3 dB 4-pole filter and acquired at a sampling frequency of 10 kHz, which is higher than the Nyquist’s critical sampling rate. Data acquisition was performed with a DigiData 1322A interface and pClamp software version 9.0 (Axon Instruments Inc., Union City, CA, USA). The dissociated VTA neurons were visualized under phase-contrast optics on an inverted microscope (Diaphot 300, Nikon, Tokyo, Japan). All experiments were performed at room temperature (23˚C - 25˚C).

Neurons were continuously bathed in the external solution and drugs were dissolved at final concentration in the same solution. Drug solutions were applied via a multiple channel manifold (MLF-4; ALA Scientific Instruments, Westbury, NY, USA). Each channel of the manifold was connected to a gravity-fed reservoir with tubing (860 μm, i.d.). The output of the manifold was connected to an outflow tube (500 μm, i.d.), the tip of which was placed within 200 mm of the soma of the recorded neuron. Solutions flowed continuously through one manifold channel. Application of drug solutions was controlled by opening or closing valves connected to the reservoirs.

Following rapid removal of the brain, the tissue block containing the VTA was mounted in the Vibratome and submerged in the ice-cold cutting solution. Coronal sections (400 μm thick) were cut and the slice was placed on a mesh platform in the recording chamber. The slice was totally submerged in the ACSF maintained at a flow rate of 2 ml/min; the temperature in the recording chamber was kept at 35˚C. The ACSF was saturated with 95% O₂ and 5% CO₂ (pH = 7.4). Equilibration time of at least one hour was allowed after placement of the brain slice in the recording chamber before electrodes were placed in the tissue. Recording electrodes were placed in the VTA under visual control. Only those neurons which were anatomically located within the VTA and which conformed to the electrophysiological criteria for dopaminergic neurons [24] were studied. These criteria include broad action potentials and regular spontaneous firing frequency at 0.5 - 5 Hz.

Extracellular recording electrodes were fabricated from 1.5 mm diameter glass tubing and were filled with 0.9% NaCl. Tip resistance of the microelectrodes ranged from 3 to 8 MΩ. The Fintronics amplifier (Fintronics Inc., Orange, CT, USA) used in these recordings had a window discriminator, the output of which was fed to both a rectilinear pen recorder and a computer-based data acquisition system for on-line and off-line analysis of the data. The multiplexed output of the amplifier was displayed on an analog storage oscilloscope, for accurate adjustment of the window levels used to monitor single units. An IBM-PC-based data acquisition system was used to calculate, display and store the frequency of firing over 5 sec and 1 min intervals.

Drugs were added to the ACSF by means of a calibrated infusion pump from stock solutions 100 to 1000 times the desired final concentrations. The addition of drug solutions to the ACSF was performed in such a way as to permit the drug solution to mix completely with the ACSF before this mixture reached the recording chamber. The use of a calibrated, variable speed infusion pump permits the accurate addition of several concentrations of drugs from the same stock solution. Final concentrations were calculated from the ACSF flow rate, pump infusion speed and the concentration of drug stock solution. The small volume chamber (about 300 μl) used in this study permitted the rapid application and washout of drug solutions. Typically drugs reached equilibrium in the tissue after 2 to 3 minutes of application.

BaCl₂, CdCl₂, CoCl₂, HEPES, NiCl₂, nystatin and ZnCl₂ were purchased from Sigma (Saint Louis, MO, USA). Papain was purchased from Worthington (Lakewood, NJ, USA).

Action potentials were analyzed offline with pClamp 9.0 software (Axon Instruments Inc.). All average values are expressed as mean ± standard error of the mean (SE). Graphing and curve fitting of data was performed with Origin 7 software (OriginLab Corp., Northampton, MA, USA). Concentration-response curves for zinc were constructed by plotting percent inhibition of I_M as a function of drug concentration plotted on a log scale. Smooth curves were fit to these data with the Hill equation of the form [25] :

where x is the concentration, y is the percent inhibition and y_max is the maximal value of y (at saturation); in the fitting procedure y_max was constrained not to exceed 100%. The term k is the IC₅₀ (the concentration giving half- maximal inhibition) and n (Hill slope) is the power term related to the slope of the curve. To assess the changes in spontaneous firing with drugs, drug effect was quantitated as the mean change in firing rate (normalized as the percentage of control) for 60 sec-long interval during the peak of the drug response as previously described [26] . The formula for this normalization is:

Data from dissociated cells with action potential amplitudes less than 50 mV were discarded. All average values are expressed as mean ± standard error (SE). Statistical comparison to assess significant differences was done by one-way ANOVA as appropriate followed by a Bonferroni correction. When needed, the Student-New- man-Keuls post hoc test was used to test multiple comparisons. Correlation was evaluated by linear regression with P < 0.05 being considered significant.

After obtaining a stable perforated patch recording, DA VTA neurons were identified in current-clamp recording based on the electrophysiological characteristics by matching spontaneous firing frequency and action potential (AP) parameters, as described previously [22] . Then, I_M was measured in the same DA VTA neuron in voltage- clamp configuration.

I_M was measured in the standard deactivation protocol [8] with 1 sec-long hyperpolarizing voltage step from a holding potential (V_H) of ?25 mV to ?40 mV. I_M was measured as the inward relaxation current caused by deactivation of I_M during the voltage step (Figure 1(a)). Figure 1(a1) shows I_M before, during and after treatment with 100 mM Zn²⁺ in a typical DA VTA neuron. Zn²⁺ reduced I_M amplitude and also reduced the sustained outward current present at ?25 mV as indicated by the inward shift in the baseline current. Just after the termination

of the hyperpolarizing voltage step from ?25 mV to ?40 mV, a transient outward current was recorded in the presence of Zn²⁺ (Figure 1(a1), middle panel). This current is likely to be transient A-type K⁺ current (I_A), because DA VTA neurons exhibit prominent I_A [27] and Zn²⁺ shifts the voltage-dependency of steady-state I_A inactivation to the depolarizing direction [28] . In the presence of zinc, I_A can be activated at a membrane potential of ?40 mV. Figure 1(a2) shows the measurement of the inward relaxation current caused by deactivation of I_M during the voltage step; the difference between the instantaneous current at the beginning (I_in) and the steady- state current at the end of the voltage step (I_ss). Figure 1(b) shows the average time course of normalized I_M amplitude before, during and after application of 100 μM Zn²⁺ in DA VTA neurons. I_M recovered completely after the washout of Zn²⁺ in DA VTA neurons.

Figure 2(a) shows that Zn²⁺ caused a concentration-dependent reduction of I_M amplitude in a typical DA VTA neuron. Zn²⁺ also caused a concentration-dependent reduction of the baseline sustained outward current. The 300 μM concentration appeared to produce a inhibitory effect near to or at the maximum, since it did not inhibit I_M substantially more than the 100 mM concentration. Zn²⁺ at 300 μM did not produce complete inhibition of I_M. Figure 2(b) shows the pooled concentration-response curve which plots normalized I_M amplitude versus log concentration of Zn²⁺ from DA VTA neurons. The Hill equation was used to fit a smooth curve to the mean data in Figure 2(b). The IC₅₀ for Zn²⁺ was 5.8 mM and the Hill slope was 0.8.

We then examined whether zinc inhibition of I_M was voltage-dependent. Figure 3(a) shows I_M induced by a series of hyperpolarizing voltage steps before, during, and after treatment with 100 mM Zn²⁺. Zn²⁺ inhibited I_M amplitude measured with all four hyperpolarizing voltage steps in 6 DA VTA neurons. I_A activation can be seen after the offset of the larger voltage steps in control and washout, while I_A is not prominent after the offset of the

larger voltage steps in the presence of Zn²⁺ (Figure 3(a)). This is likely due to a Zn²⁺-induced shift in the voltage-dependency of steady-state I_A activation in the depolarizing direction [29] . Figure 3(b) shows the relationship between membrane voltage and I_M amplitude before, during, and after treatment with 100 mM Zn²⁺ for DA VTA neurons. Zn²⁺ inhibited I_M amplitude at any voltage examined. There was no correlation between membrane voltage and the % I_M inhibition by Zn²⁺ (Figure 3(c)).

The effect of Zn²⁺ on the spontaneous firing frequency of DA VTA neurons was measured with extracellular single-unit recording of these neurons in brain slices. The percentage increase in firing frequency produced by Zn²⁺ was calculated. Figure 4 shows the concentration-response relationship between Zn²⁺ and the firing rate change by Zn²⁺ from DA VTA neurons. Average Zn²⁺-induced increase in firing frequency was 17.3% ± 10.6% with 10 mM Zn²⁺ and 49.5% ± 7.6% with 50 mM Zn²⁺ (n = 5).

Finally, we examined whether divalent cations other than Zn²⁺ modulated I_M (Figure 5). Ba²⁺, Co²⁺, Ni²⁺ and Cd²⁺ at a concentration of 100 mM all reduced I_M amplitude and induced an inward shift of the baseline outward current in DA VTA neurons, however their potencies for I_M inhibition were different (Figures 5(a)-(d)). Just after the termination of the hyperpolarizing voltage step from −25 mV to −40 mV, I_A was recorded in the pres-

ence of Cd²⁺ (Figure 5(d), middle panel), because Cd²⁺, like Zn²⁺, shifts the voltage-dependency of steady-state I_A inactivation in the depolarizing direction [28] . Figure 6(a) shows the average time course of normalized I_M amplitude before, during and after application of 100 mM divalent cations in DA VTA neurons. I_M of DA VTA neurons were inhibited by 100 mM extracellular divalent cations in increasing order of potency: Ba²⁺ < Co²⁺ < Ni²⁺ < Cd²⁺ < Zn²⁺. Average maximal inhibition of I_M was 15.6% ± 3.0 % with Ba²⁺ (n = 7), 24.9% ± 2.2 % with Co²⁺ (n = 6), 39.5% ± 4.2% with Ni²⁺ (n = 7), 59.0% ± 3.5% with Cd²⁺ (n = 6), and 66.8% ± 2.8% with Zn²⁺ (n = 6).