Amphetamine Regulation of Dopamine Transport

Dopaminergic neurotransmission is fine-tuned by the rate of removal of dopamine (DA) from the extracellular space via the Na+/Cl--dependent DA transporter (DAT). DAT is a target of psychostimulants such as amphetamine (AMPH) and cocaine. Previously, we reported that AMPH redistributes the human DAT away from the cell surface. This process was associated with a reduction in transport capacity. This loss of transport capacity may result either from a modification of the function of DAT that is independent of its cell surface redistribution and/or from a reduction in the number of active transporters at the plasma membrane that results from DAT trafficking. To discriminate between these possibilities, we stably transfected HEK-293 cells with a yellow fluorescent protein (YFP)-tagged human DAT (hDAT cells). In hDAT cells, acute exposure to AMPH induced a time-dependent loss of hDAT activity. By coupling confocal imaging with patch-clamp whole-cell recordings, we have demonstrated for the first time that the loss of AMPH-induced hDAT activity temporally parallels the accumulation of intracellular hDAT. In addition, presteady-state current analysis revealed a cocaine-sensitive, voltage-dependent capacitance current that correlated with the level of transporter membrane expression and in turn served to monitor the AMPH-induced trafficking of hDAT. We found that the decrease in hDAT cell surface expression induced by AMPH was not paralleled by changes in the ability of the single transporter to carry charges. Quasi-stationary noise analysis of the AMPH-induced hDAT currents revealed that the unitary transporter current remained unaltered during the loss of hDAT membrane expression. Taken together, these data strongly suggest that the AMPH-induced reduction of hDAT transport capacity results from the removal of active hDAT from the plasma membrane.

The Na ϩ /Cl Ϫ -dependent transporter family includes plasmalemmal carriers for monoamines such as DA, 1 serotonin, and norepinephrine (1,2). DAT tunes the spatial and temporal characteristics of dopaminergic neurotransmission by regulating extracellular DA concentration (3,4). Although diffusion and enzymatic degradation also reduce the synaptic concentration of this monoamine, the development of a DAT knockout mouse established reuptake as the primary mechanism controlling extracellular DA levels (3).
Dopaminergic neurotransmission mediates numerous biological events, including reward, addiction, movement, and lactation (1,5). Several therapeutic agents (6), environmental toxins (7,8), and psychostimulants (AMPH and cocaine) (9) have been shown to generate their effects by targeting DAT. AMPH rapidly decreases DA clearance and stimulates an DATmediated DA efflux (10). The subsequent increase in dopaminergic signaling in limbic areas of the brain is believed to mediate the rewarding and addictive properties of AMPH (9). In 2000, Saunders and colleagues proposed a novel action of AMPH by demonstrating that acute application of AMPH reduces hDAT cell surface expression (11). Sorkina and colleagues (12) recently extended this observation by showing that AMPH induced an intracellular accumulation of hDAT in early and late endosomal vesicles, which coexpress the endosomal proteins Rab5, Rab11, Hrs, and EEA.1.
Regulators of DAT activity include AMPH and cocaine (11)(12)(13)(14)(15)(16), G-protein-coupled receptors such as the D 2 and mGluR5 receptors (17,18), kinases such as protein kinase C, phosphatidylinositol 3-kinase, tyrosine kinase, and Ca 2ϩ /calmodulin kinase (12, 13, 19 -23), and transporter interacting proteins (24,25). In addition, some of these studies suggest that the regulation of DAT activity may originate from a change in DAT cell surface expression (11, 13-17, 19, 21, 23-25). Although these results suggest that trafficking of the transporter modulates the transport capacity of the system, DAT could be modified and functionally inactivated by AMPH prior to trafficking from the plasma membrane. Therefore, the extent to which hDAT trafficking determines the AMPH-induced down-regulation of hDAT activity is unknown. The present work combines confocal imaging, whole cell steady-state and transient current recordings, as well as noise analysis, to simultaneously monitor DAT cell surface expression and activity. Our data suggest that upon AMPH application hDAT redistributes from the plasma membrane as an active carrier.

MATERIALS AND METHODS
Cell Culture-A fluorescently tagged hDAT was constructed by fusing the C terminus of the coding region of enhanced yellow fluorescent protein (YFP) from pEYFP-N1 (Clontech) to the N terminus of the human synthetic DAT cDNA, thereby creating the fusion construct YFP-hDAT. This construct was subcloned into a bicistronic expression vector (26) modified to express the synthetic hDAT from a cytomegalovirus promoter and the hygromycin resistance gene from an internal ribosomal entry site (pciHyg), as described previously (11). EM4 cells, an HEK 293 cell line stably transfected with macrophage scavenger receptor (27) (R. Horlick, Pharmacopeia, Cranberry, NJ), were transfected with the YFP-DAT using LipofectAMINE (Invitrogen), and a stably transfected pool (hDAT cells) was selected in 250 g/ml hygromycin as described previously (11,28). Cells were grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum at 37°C and 5% CO 2 . Reagents were purchased from Sigma (St. Louis, MO) unless otherwise noted.
Electrophysiology and Confocal Microscopy-Parental or stably transfected cells were plated at a density of 10 5 per 35-mm culture dish. Before electrical recordings, attached cells were washed twice with the bath solution containing the following (in mM): 130 NaCl, 10 HEPES, 1.5 CaCl 2 , 0.5 MgSO 4 , 1.3 KH 2 PO 4 , and 34 dextrose adjusted to pH 7. 35. In experiments investigating the Na ϩ dependence of the hDAT transient charge movement, Na ϩ was iso-osmotically replaced with choline. The hDAT cells were sequentially perfused with bath solutions containing the following Na ϩ concentrations (in mM): 3, 24, 53, 94, 122, and 130. The recording pipette was filled with a solution containing the following (in mM); 120 KCl, 2.0 MgCl 2 , 0.1 CaCl 2 , 1.1 EGTA, 10 HEPES, and 30 dextrose adjusted to pH 7.35. Free Ca ϩ2 was calculated to be 0.1 mM. A programmable puller (model P-2000, Sutter Instruments, Novato, CA) was used to fabricate quartz-recording pipettes with a resistance of 5 M⍀. Whole cell currents were recorded using an Axopatch 200B with a low pass Bessel filter set at 5,000 Hz. Current-voltage relationships were generated by stepping the membrane voltage from a holding potential of Ϫ20 mV to voltages between Ϫ160 and 100 mV in 20-mV increments for 1 s. A waveform generator (Challenger VM-2C) was used to vary the membrane potential. Data were stored on a VCR and analyzed with a Nicolet Integra Model 20 oscilloscope and a DELL computer using programs written by W. N. Goolsby (Emory University, Atlanta, available on request).
Application of dextro-AMPH to hDAT cells generated an hDAT current, which after several minutes of stability began to decrease. Because of cell-to-cell variability, we defined "time zero" as the onset of the decreasing phase of the AMPH-induced current. We subsequently defined the initial hDAT current trace (time zero) as the AMPH-induced current recorded immediately before the onset of the decreasing phase of the hDAT-mediated current. The final current was defined as the last current trace recorded before adding cocaine. The AMPH-induced, hDAT-mediated steady-state current and transient charge movement was obtained by subtracting the current trace recorded in the presence of AMPH plus cocaine from the current trace recorded in the presence of AMPH.
The hDAT transient charge (Q) movement in response to a voltage step, was obtained by integrating either the "on" or the "off" of the relaxation component of the AMPH-induced current ( Fig. 2A). Time-dependent changes in Q were used to evaluate hDAT cell surface expression as described previously (29 -36). The charge-voltage relationships were fit to the Boltzmann function, (Q dep and Q hyp are the charge movements at the depolarizing and hyperpolarizing limits), V is the membrane voltage, V1 ⁄2 is the potential at which half of the charge has moved, q is the elementary charge, z␦ is the product of the valence of the moving charge and the fraction of the membrane field through which it moves, k is the Boltzmann constant, and T is the absolute temperature. For simplicity, the slope factor of the fitting was kT/qz␦. Charge-voltage relationships from each experiment were fit, and the results were reported as mean Ϯ S.E.
Unless otherwise noted, the steady-state current (I) at a particular voltage was calculated as the average current during the final 100 ms of the voltage step. The time course for the AMPH-induced decrease in hDAT steady-state current (Fig. 5A) was fit with the Boltzmann equation, I ϭ (A 1 Ϫ A 2 )/(1 ϩ exp(t Ϫ t 50 )/␣) ϩ A 2 ), where A 1 is the initial current, A 2 is the final current, t 50 is the time at which half of A 1 remains, and ␣ is the slope of the fit. For all the experiments, the t 50 time was reported as mean Ϯ S.E.
To synchronize the acquisition of hDAT-mediated currents with the confocal imaging of the cellular distribution of hDAT, we simultaneously triggered the voltage steps and the confocal imaging software (EZ 2000, Coord Automatisering, The Netherlands) with the Challenger VM-2C. Cells were visualized with a Nikon eclipse TE300 inverted microscope. Images were acquired using 488-nm excitation with a 510-nm long pass filter. The image acquired during the initial hDAT current trace (time zero as described above) was defined as the initial confocal image. To minimize the fluorophore bleaching and cytotoxic effects of confocal microscopy, images were acquired every 1.5 min prior to the start of the hDAT current decreasing phase and every 30 s during the decreasing phase. Because membrane proteins such as GLUT4 can be localized to the plasma membrane without being exposed to the extracellular space (37), we quantified intracellular instead of plasma membrane fluorescence as a measure of AMPH-induced hDAT membrane redistribution. Therefore, by excluding plasma membrane fluorescence, we may underestimate the amount of intracellular DAT. The quantitation of intracellular fluorescence was performed by normalizing for fluorophore bleaching and cytosolic volume changes (38,39). ImageJ software (National Institutes of Health, available online) was utilized for this analysis. Normalized data were analyzed with Prism 3.02 software (GraphPad Software, Inc.) and reported as mean Ϯ S.E. unless otherwise indicated. The reported correlation coefficient (r) was obtained by linear regression analysis. Uptake of [ 3 H]DA-[ 3 H]DA uptake was performed as previously described (11,13). hDAT cells were seeded into 24-well plates ϳ24 h prior to the experiment (150,000 cells per well). After 5 h of serum starvation, the cells were treated in quadruplicate wells with AMPH in uptake buffer, containing (in mM); 120 NaCl, 4.7 KCl, 10 HEPES, 5 Trizma base, 2.2 CaCl 2 , and 10 dextrose with 100 M ascorbic acid at pH 7.4 and 37°C. The plates were removed from the incubator, and the cells were washed (three washes of 5 min each) with 4°C uptake buffer to remove the AMPH from each well and inhibit protein trafficking. The plates were then placed into an 18°C incubator in uptake buffer containing 100 M pargyline, a monoamine oxidase inhibitor. Then 50 nM [ 3 H]DA (PerkinElmer Life Sciences, Boston, MA) together with 15 m DA was added to reach a final volume of 250 l. Cells were incubated for 2 min and then the solution was aspirated to terminate uptake. After three quick washes with ice-cold uptake buffer, the cells were lysed with 300 l of 1% SDS. Radioactivity was measured in a Beckman scintillation counter with UniverSol mixture. Specific uptake was defined as total uptake minus nonspecific uptake in the presence of 10 m mazindol. Data were analyzed with Prism 3.02 software and reported as mean Ϯ S.E.
Cell Surface Biotinylation-Cell surface biotinylation experiments were performed as previously described (40) with slight modification. hDAT cells were seeded into 6-well plates (10 6 cells/well) ϳ48 h prior to the experiment. After 1 h of serum starvation, the cells were washed twice with 37°C uptake buffer and treated with AMPH in uptake buffer at 37°C. The cells were then washed twice with ice-cold PBS containing 0.1 mM CaCl 2 and 1 mM MgCl 2 (PBS-Ca-Mg) and treated with Sulfo-NHS-S-S-Biotin (1.5 mg/ml in PBS-Ca-Mg, Pierce Chemical Co., Rockford, IL) on ice for 1 h. The reaction was quenched by washing twice with 4°C PBS-Ca-Mg containing 100 mM glycine (PBS-Ca-Mg-glycine) followed by an incubation with PBS-Ca-Mg-glycine for 30 min on ice. Cells were then washed twice with 4°C PBS-Ca-Mg before lysis with 1 ml of radioimmune precipitation assay buffer (20 mM Tris, 20 mM EGTA, 1 mM dithiothreitol, 1 mM benzamidine, 1% Triton X-100: supplemented with protease inhibitors 100 M phenylmethylsulfonyl fluoride, 5 g/ml leupeptin, 5 g/ml pepstatin) for 30 min on ice with constant shaking. Lysates were centrifuged at 14,000 ϫ g for 30 min at 4°C. The supernatants were isolated and biotinylated proteins were separated by incubation with ImmunoPure Immobilized Streptavidin beads (Pierce Chemical Co., Rockford, IL) for 1 h at room temperature with constant mixing. Beads were washed three times with radioimmune precipitation assay buffer, and biotinylated proteins were eluted with Laemmli loading buffer for 30 min at room temperature. Total cell lysates and biotinylated proteins (cell surface) were separated by SDSpolyacrylamide electrophoresis (7.5%) and transferred into polyvinylidene difluoride membranes (Millipore). Polyvinylidene difluoride membranes were incubated for 1.5 h in blocking buffer (5% dry milk, 0.1% Tween 20 in Tris-buffered saline) and immunoblotted with a rat monoclonal antibody directed against the N terminus of the human dopamine transporter (1:2000 in blocking buffer, Chemicon Inc., Temecula, CA). Immunoreactive bands were visualized using horseradish peroxidase-conjugated goat anti-rat antibody (1:5000 in blocking buffer, Santa Cruz Biotechnology, Santa Cruz, CA) with ELC-Plus on Hypersensitive ECL film (Amersham Biosciences, Arlington Heights, IL). Band densities were calculated using Scion-Image software (Scion Corp., Frederick, MD) and normalized to the appropriate total extract to control for protein loading. Data were analyzed with Prism 3.02 software and reported as mean Ϯ S.E.

RESULTS
The human EM4 cell line provides an appropriate parental background for studying hDAT function because of the absence of [ 3 H]DA uptake and the lack of either DA-or AMPH-induced whole cell currents (11). To investigate the extent to which transporter trafficking regulates transport capacity, we created a pool of cells stably expressing fluorescently tagged hDAT. Using confocal imaging, cells with easily visualized plasma membrane fluorescence were selected for analysis ( Fig.  1A, inset). Addition of the N-terminal YFP tag to DAT did not significantly alter [ 3 H]DA uptake (data not shown) and did not perturb the ability of the transporter to produce substrateinduced currents (Fig. 1A) (11,41). Fig. 1A shows current traces recorded from an hDAT cell before substrate application (CTR), after bath application of 10 M AMPH (AMPH), and after the addition of 10 M cocaine with AMPH still present (COC). The AMPH-induced hDAT-mediated current was calculated as the current recorded upon bath application of AMPH minus the current recorded after the addition of cocaine, with AMPH still present. Cells without detectable fluorescence did not produce any measurable hDAT-mediated whole cell current upon AMPH application (data not shown). Fig. 1B shows the current-voltage relationship for the AMPH-induced current. The membrane voltage of the cell was held at Ϫ20 mV and stepped every 4 s, in 20-mV increments, to voltages between Ϫ160 mV and ϩ100 mV. Data were normalized to the current measured at Ϫ160 mV (Ϫ60.2 Ϯ 16.9 pA, n ϭ 8). The AMPHinduced current reversed at potentials more positive than Ϫ35.7 Ϯ 4.0 mV (Fig. 1B).
Cocaine application induced a minor decrease of the control current due to its ability to block the DAT-mediated leak conductance (Fig. 1A). This substrate-independent current, which has been described previously for DAT (41,42) as well as other neurotransmitter transporters (30,31) constitutes, in HEK 293 cells expressing hDAT, only a small percentage of the total AMPH-induced current at negative voltages (11,43). Moreover, DAT substrates (such as DA and AMPH) have also been shown to strongly inhibit this leak current (41,42). In hDAT cells the leak current revealed by bath application of 10 M cocaine was on average Ϫ7.4 Ϯ 1.6 pA at Ϫ160 mV (n ϭ 8), ϳ10% of the AMPH-induced steady-state current.
Following a voltage jump, hDAT cells display current relaxations that persist for several milliseconds (Fig. 1A) after the time required to charge the membrane capacitance. The hDAT component of these current relaxations (hDAT-mediated transient current) can be defined by bath application of DA uptake blockers such as mazindol (11) and cocaine ( Fig. 1A). As for the norepinephrine transporter (31), the hDAT-mediated transient current was approximately the same in the presence or absence of substrate (e.g. AMPH, data not shown). hDAT-mediated transient currents were isolated by subtracting the current recorded in the presence of AMPH and cocaine from the current recorded in the presence of AMPH. Fig. 2A shows a representative hDAT-mediated transient current obtained by stepping the membrane voltage from Ϫ20 to Ϫ120 mV. hDAT-mediated transient currents were not seen in untransfected cells (NT) (n ϭ 5, Fig. 2B). The charge movement (Q) in response to a voltage jump was calculated for a range of test potentials by integrating the transient currents (shaded area, Fig. 2A). At Ϫ160 mV, the magnitude of the transient charge movement, that is close to saturation, correlates with the level of transporter activity as measured by AMPH-induced steady-state currents (r ϭ 0.88, n ϭ 12, Fig. 2C). Additionally, AMPH induced a time-dependent decrease of transporter membrane expression as measured by cell-surface biotinylation (Fig. 2D). This decrease temporally correlated with the decrease in hDAT transient currents obtained in parallel experiments (Fig. 2D). At Ϫ160 mV the AMPH-induced time-dependent decrease in hDAT-mediated transient current was still present after extensively washing away AMPH for 1 min (data not shown), showing that the decrease in transient current reflects cell surface expression. By plotting the average values of Fig. 2D against each other, we verified that a strong linear correlation exists between the decrease in hDAT surface expression measured by biotinylation experiments and the decrease in hDAT transient currents (r ϭ 0.90, graph not shown). In addition, the hDATmediated transient currents were dependent on extracellular Na ϩ concentration, as were the steady-state currents. The hDAT-mediated transient current obtained by stepping the membrane potential to Ϫ160 mV from a holding potential of Ϫ20 mV increased linearly as the extracellular Na ϩ concentration was increased from 3 to 130 mM (data not shown). The slope of the linear regression obtained by plotting the transient charge movement obtained against Na ϩ concentrations was Ϫ0.0049 Ϯ 0.0004 and r ϭ 0.99. Taken together, these data indicate that, as shown previously for other transporters (29 -36), DAT transient charge movements can be used to evaluate relative changes in DAT surface expression.
Previous reports have utilized a fluorescent protein-tagged DAT to relate alterations in DAT activity with trafficking (13,20). Fig. 3A shows the cellular distribution of DAT in hDAT cells under basal conditions (Basal). Ten minutes after perfus- ing non-serum-starved hDAT cells with 10 M AMPH, an increase of intracellular fluorescence was detected. During the 60 min of AMPH exposure, the intracellular fluorescence increased to 131.25 Ϯ 0.03% of the basal value (Fig. 3B). Perfusion of vehicle alone did not alter the cellular distribution of the transporter (intracellular fluorescence was 99.89 Ϯ 0.02% of the basal value after 60 min).
Treatment of non-serum-starved hDAT cells with 10 M AMPH produced a time-dependent decrease in [ 3 H]DA uptake (Fig. 3C). After a 60-min treatment, [ 3 H]DA uptake was 75.5 Ϯ 0.2% of control (n ϭ 4). To determine the temporal relationship between the AMPH-induced hDAT cell surface redistribution and the reduction of hDAT activity, we plotted the percent increase of intracellular fluorescence (Fig. 3B) against the percent inhibition of DA uptake (Fig. 3C) for each experimental time point. Fig. 3D shows a significant correlation (r ϭ 0.75) between redistribution of hDAT away from the cell surface and the inhibition of [ 3 H]DA uptake after exposure to AMPH.
To correlate AMPH-induced changes in hDAT-mediated ionic flux with changes in hDAT localization in individual cells, we used the patch-clamp technique in the whole cell configuration. After bath application of AMPH, and after several minutes of stability the AMPH-induced current began to decrease. As described under "Materials and Methods," we defined the initial hDAT current trace (time zero) as the AMPH-induced current recorded immediately before the onset of the decreasing phase, and the final current was defined as the last current trace recorded. On average, the "initial" current trace was recorded 10.3 Ϯ 0.5 min after AMPH bath application (see "Materials and Methods"). We normalized both the steadystate and transient currents recorded during each experiment to their initial values acquired at Ϫ160 mV (Ϫ95.9 Ϯ 23.4 pA and Ϫ10.9 Ϯ 4.4 pC, respectively; n ϭ 5). Fig. 4A, shows the current-voltage relationship for the initial and final AMPH- induced currents. The average decrease in AMPH-induced steady-state current was 62.5 Ϯ 3.7% (for each individual voltage, the initial current value was compared with the final value (n ϭ 5), and these values were averaged for every voltage tested). No significant shift in the reversal potential of the hDAT currents was observed (from Ϫ36.1 Ϯ 4.0 mV to Ϫ45.7 Ϯ 9.9 mV, initial values compared with final values) confirming that the loss of hDAT activity was not due to accumulation of intracellular AMPH or a change in the ionic gradients driving hDAT activity.
Exposure to AMPH also reduced hDAT cell surface expression as measured by transient current analysis. The transient current for both the "on" and "off " components of each voltage step was integrated to reveal the total charge movement in response to the voltage step (Fig. 4B). It should be noted that the hDAT-mediated transient currents, like those for the ␥-aminobutyric acid transporter GAT-1 (29), proton/myo-inositol cotransporter (32), and the Na ϩ /glucose cotransporters SGLT1 (34) and SGLT2 (33), appear to be capacitive in nature, because the "on" and "off " charge movements are equal and opposite in sign. The transient charge decreased an average of 67.3 Ϯ 1.5% for the "on" and 69.9 Ϯ 3.7% for the "off " for each voltage tested (initial values compared with final values, n ϭ 5). Similarly, after 10 min of treatment with 1 M PMA the hDAT-mediated transient current was decreased 65 Ϯ 10% at Ϫ160 mV (n ϭ 5). Although the PMA effect was not further characterized, the reduction in transient currents is consistent with previous findings that PMA also causes a cell surface redistribution of DAT (Loder and Melikian (23)), further supporting the use of transient currents to quantitate cell surface DAT.
To simultaneously measure hDAT cell surface expression and transporter activity, we analyzed the transient and steadystate currents obtained from an individual voltage step (Ϫ160 mV). Beginning with the initial current trace, the values for the transient charge movement (Fig. 4C) and steady-state current (Fig. 4D) were plotted against time. The loss of hDAT surface expression measured by transient current analysis mirrored the reduction of hDAT activity. The reduction in transient current in response to AMPH temporally correlated with the decrease of steady-state current (r ϭ 0.87) (Fig. 4E), suggesting that the AMPH-induced hDAT trafficking directly regulates hDAT activity.
The hDAT-mediated transient charge movement depended on the membrane potential (Fig. 4B, filled squares) in a manner well described by a Boltzmann function (see "Materials and Methods"). For "initial" (time at which we recorded the initial current), the total charge movement (Q max ) in cells expressing hDAT was on average 16.4 Ϯ 6.5 pC, the voltage midpoint of the charge movement (V1 ⁄2 ) was Ϫ71.8 Ϯ 16.9 mV, and the slope factor of the function was 40.8 Ϯ 7.0 mV. Although Q max decreased 66.5 Ϯ 9.0%, no significant changes were found for the V1 ⁄2 (Ϫ66.3 Ϯ 13.8 mV) or the slope factor (31.8 Ϯ 3.1 mV) over the course of the experiment. Accordingly, the initial value for z␦ (calculated from the slope factor) was 0.69 Ϯ 0.1, whereas its final value was 0.84 Ϯ 0.1. These data are consistent with several models in which one charge or a combination of many positive and/or negative charges moves, resulting in a z␦ of ϳ0.8 (29,32,36). Because z␦ and V1 ⁄2 were stable during the entire duration of the experiments, these data are consistent with the changes in Q max upon AMPH application being due solely to changes in the number of hDAT at the plasma membrane. If one elementary charge is translocated per molecule of neurotransmitter, then the turnover rate can be directly calculated from the transporter density (Q max ) and the steady-state currents. Formally, the turnover rate is equal to (I/Q max )*z␦, where I is the steady-state current and Q max and z␦ are defined as above (29,32). For the dopamine transporter, not all the substrate-induced current is stoichiometrically coupled to substrate translocation (42). However, for a specific voltage the ratio of charge to translocated substrate molecule is constant (41,42). Given that z␦ was not significantly altered from the beginning to the end of the experiment, we employ the ratio I/Q max to monitor transporter activity over time. At a potential of Ϫ160 mV, upon 10 M AMPH application the initial value of I/Q max was 7.4 Ϯ 3.1 s Ϫ1 , whereas its final value was 6.5 Ϯ 2.7 s Ϫ1 (n ϭ 5). Therefore, despite a decrease of the number of transporters on the cell surface, as reflected by the 67% decrease in Q max , the intrinsic activity of each measurable hDAT molecule was unaltered.
To further test whether AMPH regulates hDAT activity solely by inducing hDAT trafficking, we coupled confocal imaging with the patch clamp technique in the whole cell configuration. Intracellular fluorescence, transient charge movement, and steady-state current were measured simultaneously every 30 s by acquiring a confocal image during each voltage step. Fig. 5A shows the initial hDAT cellular distribution and the hDAT current after bath application of 10 M AMPH (Initial).
The initial current traces and confocal images were defined as described under "Materials and Methods." Eight minutes after acquiring the initial trace/image, an increase of intracellular fluorescence and a concomitant reduction in the hDAT steadystate and transient currents were observed (Fig. 5B). The average time that elapsed between bath application of AMPH and the initial current trace/image was 10.2 Ϯ 1.5 min (n ϭ 4). We plotted the decrease in steady-state and transient currents induced by AMPH against the increase of hDAT intracellular fluorescence for each of the experiments (Fig. 5, C and D). It's worth noting that the analysis of intracellular fluorescence (excluding membrane fluorescence, see "Materials and Methods") may underestimate the level of intracellular DAT, because membrane proteins can be localized to the plasma membrane with confocal microscopy without being exposed to the extracellular space (37). The time-dependent reduction in hDAT steady-state currents was temporally correlated with the increase in intracellular fluorescence (r ϭ Ϫ0.72, Fig. 5C). In addition, a strong correlation was observed between the reduction of hDAT surface expression measured by transient currents and the increase in intracellular transporter measured by confocal imaging (r ϭ Ϫ0.68) (Fig. 5D). Changes in the cellular fluorescence distribution or hDAT-mediated currents were not observed in cells that were treated with vehicle alone (data not shown, n ϭ 3).
If trafficking of hDAT and not DAT modification controls transport capacity in response to AMPH, then each individual transporter would be expected to function normally until removed from the plasma membrane. In the whole cell configuration, we recorded the AMPH-induced current at Ϫ80 mV and analyzed the current fluctuations with quasi-stationary noise analysis. To facilitate analysis and provide a high sampling resolution, the raw data traces were examined in blocks of 500 ms. Fig. 6A shows the decreasing phase of the AMPH-induced current over time for a representative experiment. As described under "Materials and Methods," time zero was defined as the beginning of the descending phase of the current (arrow) (10.1 Ϯ 2.4 min following AMPH bath application (n ϭ 5)), and data preceding time zero were partially omitted to highlight the decreasing phase. For each experiment, the decreasing phase was fit to the Boltzmann equation (see "Materials and Methods"). The average t 50 for the five experiments was 3.7 Ϯ 0.64 min, whereas ␣ was 1.24 Ϯ 0.37 min. The variance of the AMPH-induced current was defined as the variance calculated in the presence of AMPH minus the variance calculated in the presence of 10 M AMPH and 10 M cocaine. Although bath application of cocaine reduced the whole cell current variance recorded in the presence of AMPH by 42.6 Ϯ 4.1%, cocaine decreased the current variance recorded in control condition by only 2.7 Ϯ 1.0%. These data, combined with the fact that the leak current represents only 10% of the AMPH-induced current (see above), suggest that the contribution of the leak conductance only minimally affects our analysis of the variance of the hDAT current. Fig. 6B shows that the variance of the AMPHinduced current (Fig. 6A) decreased over time in parallel with the decrease of the AMPH-induced current. If we assume that hDAT behaves similarly to an ion channel (42,43), then the relationship between steady-state current and the variance is given by the equation: 2 /I ϭ i(1 Ϫ p); where 2 is the variance, I is the steady-state current, i is unitary transporter current, and p is the probability of opening of the channel. The ratio of variance to current was constant during the AMPH-induced decreasing phase of the hDAT whole cell current (Fig. 6C, i(1 Ϫ p) ϭ Ϫ0.034 Ϯ 0.0089 pA), demonstrating that, even though the macroscopic transporter activity was reduced, the microscopic properties of the individual transporter were unchanged.
By stepping the membrane voltage to Ϫ80 mV from a holding potential of Ϫ20 mV, for 1 s, every 500 ms, we were able to correlate at a higher time resolution the loss of hDAT cell surface expression (as reflected by the transient current, Q) with the reduction in hDAT activity (as reflected by the AMPHinduced current, I), while simultaneously monitoring with quasi-stationary noise analysis the biophysical characteristics of the individual transporter. hDAT surface expression and activity decreased simultaneously (r ϭ 0.67 Ϯ 0.07, mean Ϯ S.E., n ϭ 4) throughout the 45 s of measurement during the decreasing phase of hDAT activity. The first step of 30 consecutive voltage steps (step 1) was 16.1 Ϯ 2.4 min following AMPH bath application (Fig. 7A). In contrast, we observed no correlation between the decrease in hDAT cell surface expression and unitary transporter properties, i(1 Ϫ p), measured by the ratio 2 /I (Fig. 7B, r ϭ 0.16 Ϯ 0.06, mean Ϯ S.E., n ϭ 4). These data strongly suggest that the individual transporter protein is active and functions normally until removed from the plasma membrane in response to acute AMPH exposure. DISCUSSION Fine-tuning the clearance of neurotransmitter to the acute requirements of synaptic transmission is a vital aspect of neuronal plasticity. Both the number of active transporters at the plasma membrane and the rate of transport of an individual transporter regulate the transport capacity of a cell. The reduction in DA uptake induced by 1 h of exposure to the DAT substrate AMPH was associated with an increase of intracellular hDAT (11). In addition, Saunders and colleges were able to block the AMPH-induced down-regulation of hDAT cell surface expression and the associated decrease of DA uptake by overexpression of a dominant negative mutant of dynamin 1 (K44A) a GTPase known to play a role in clathrin-mediated endocytosis (44,45), caveolae endocytosis (46), and the activation of mitogen-activated protein kinase (47,48).
Because AMPH has been shown to activate enzymes such as extracellular signal-regulated kinase 1/2 (49) and protein kinase C (50), it is possible that AMPH stimulation of these signaling pathways directly regulates hDAT activity independent of trafficking. In such a scenario, hDAT cell surface redistribution could represent a secondary mechanism by which cells control the number of modified hDAT on the cell surface. Thus, we sought to address whether AMPH regulates hDAT activity while the transporter is at the plasma membrane or whether AMPH modulates hDAT function principally by redistributing active carriers from the plasma membrane. Our data demonstrate that AMPH regulates DA transport capacity directly through a trafficking-dependent mechanism.
The loss of DA uptake caused by AMPH was associated with an increase in intracellular hDAT. This is consistent with other reports showing that cocaine (14,15), receptor activation (17,51), signal transduction pathways (13,19,21), and cytosolic proteins (24,25) regulate DAT activity with concomitant modulation of DAT cell surface expression. However, establishing a strong correlation between cell surface redistribution of DAT and the decrease in transporter function has been restricted by the limitations of the methodologies employed. Indeed, while appropriate for measuring the average characteristics from a population of transporters, techniques such as [ 3 H]DA uptake and cell surface biotinylation are limited by a low time resolution. In addition, important cellular aspects driving uptake activity (e.g. voltage and ionic gradients (41,42)) cannot be accurately controlled with these experimental methodologies. For example, it seems likely that the reduction in [ 3 H]DA uptake induced by AMPH after 1 and 5 min (Fig. 3C) reflects a reduction in the gradients driving DA uptake and not a reduction in hDAT cell surface expression. This hypothesis is supported by the observation that AMPH is able to depolarize the plasma membrane of hDAT cells (upon amphetamine application, the cell depolarized 13 Ϯ 2 mV (n ϭ 3)) and that, under voltage clamp, the AMPH-induced reduction of hDAT activity, measured by electrical currents, takes on average 10.2 Ϯ 1.5 min to occur. Even after several min of washing away AMPH, hDAT cells were still significantly depolarized (data not shown).
The patch clamp technique in whole cell configuration simultaneously measures both the AMPH-induced transient and steady-state currents in response to a voltage step. Several lines of evidence support the conclusion that the cocaine-sensitive transient currents originate from hDAT and correlate with hDAT cell surface expression: 1) they are absent in nontransfected cells; 2) they are blocked by cocaine and mazindol, two pharmacological blockers of hDAT activity; 3) their decrease temporally correlates with a decrease in hDAT cell surface expression measured by biotinylation, steady-state currents, and intracellular fluorescence; 4) there is a tight correlation between the amount of transient charge movement and steady state current between cells; and 5) hDAT-mediated transient currents depend on Na ϩ , which is required for DA uptake. In fact, several reports have utilized transient currents to measure cell surface expression of biogenic amine transporters (29,31,36). For the norepinephrine transporter (NET), the transient charge movement was found to be independent of substrate and was blocked by desipramine, a pharmacological blocker of NET activity (31). These authors concluded that the NET transient current is the result of the movement of transporter-associated charges in the membrane field in response to the voltage step. In addition, for the GAT-1, the transient charge movement was found to be strictly dependent on extracellular Na ϩ , a cotransported ion (29). The authors interpreted the data as ions binding and unbinding the transporter and not as the result of the movement of charges within the protein structure. Regardless of the mechanistic interpretation, in both models, the magnitude of the transient charge movement is proportional to the number of transporters at the plasma membrane, and therefore a direct measure of transporter cell surface expression. Furthermore, freeze-fracture electron microscopy studies have confirmed that transient current analysis is a valid method to estimate plasma membrane expression for transporters (34).
The AMPH-induced steady-state current arises from the movement of ions into the cell during the transport cycle and, thus, is a direct measurement of hDAT electrical activity (42). By following relative changes in the transient charge movement and the steady-state current, we are able to measure, at the single cell level and with a high time resolution, both hDAT cell surface expression and activity.
The hDAT transient currents behaved as a classic capacitive charge movement (31), because the transient charge movement associated with the "on" and "off " components of the voltage step were equal in magnitude, opposite in sign, and decreased proportionally over time upon AMPH application. We found a strong correlation between the decrease in transient and steady-state currents (Fig. 4) indicating that, as the hDATmediated movement of ions decreases, there is a decrease in transporter surface expression.
In addition, we coupled confocal microscopy with whole cell patch clamp recordings. In these experiments we utilized two independent methods to follow protein redistribution (confocal imaging and transient charge movement). We demonstrated that AMPH induces a loss of hDAT-mediated transient currents, which correlates with an increase in intracellular fluorescence (Fig. 5D). More importantly, we found a strong temporal correlation between the decrease in hDAT steady-state currents and the increase in intracellular fluorescence (Fig.  5C). Therefore, these data along with previous reports in the literature support the conclusion that changes in transient currents reflect changes in hDAT cell surface expression (29 -36).
If AMPH regulates transport capacity solely by changing the number of transporters on the plasma membrane, then the activity of the individual transporter protein would be unchanged until AMPH causes its internalization. To test this, quasi-stationary noise analysis was performed on the hDATmediated steady-state current to monitor the hDAT unitary currents over time. Unlike radiolabeled uptake, which measures the average activity for a population of transporters, analysis of the current fluctuation reveals the properties of the individual active transporter. In particular, this protocol allowed the analysis of hDAT activity, at a membrane voltage (Ϫ80 mV) that supports inward AMPH-induced currents, with a higher time resolution. Any error introduced by defining the AMPH-mediated current by cocaine sensitivity is most likely negligible, because the DAT leak conductance constitutes only 10% of the total AMPH-induced current (see "Results"). Fig. 6 shows that the AMPH-induced steady-state current decreased over time in parallel with a reduction in the current fluctuations. However, the ratio 2 /I ϭ i(1 Ϫ p) remained constant throughout the experiment. As for the norepinephrine transporter (52), the probability of opening for the hDAT unitary current is likely very small, and therefore this equation would simplify to 2 /I Ϸ i, which is the upper limit. Thus, although alternative models are conceivable, these results suggest that, during the AMPH-induced decrease in hDAT electrical activity, the ability of the individual transporter to conduct charges was unchanged. If we assume that the steady-state current is given by the equation I ϭ Nip, where N is the total number of transporters and i and p are constant (Fig. 6C), then the macroscopic hDAT current must decrease due to a reduction in the number of transporters. Therefore, the reduction of transporters at the plasma membrane is likely to result from a trafficking phenomenon, although an extremely rapid inactivation of the transporter protein cannot be excluded by noise analysis alone. However, the process of down-regulating the hDAT activity induced by AMPH must be relatively slow, because the decrease over time of the AMPH-induced current (Fig. 6A) during the descending phase had a time constant (t 50 ) of 3.7 Ϯ 0.64 min. In this context, the strong temporal correlation between the decrease in transient charge movement and the FIG. 7. AMPH redistributes hDAT from the plasma membrane as an active protein. The whole cell hDAT current induced by 10 M AMPH was recorded in response to a voltage step from a holding potential of Ϫ20 to Ϫ80 mV every 500 ms. The transient charge movement (Q), steady-state current (I), and the ratio of the variance of the steady state current ( 2 ) to the mean current (I) were calculated from each AMPH-induced current trace. The reported data points are from 30 consecutive voltage steps during the decreasing phase of hDAT activity. Data were normalized to the corresponding values (Q, I, and 2 /I) calculated from the first voltage step of the sequence (step 1). Each point represents data obtained from a single voltage step. A, parallel reduction of transient charge movement Q and steady-state current I. B, the AMPH-induced reduction in transient current does not correlate with any change in the unitary transporter current as estimated by 2 /I. increase in intracellular fluorescence (Fig. 5D) demonstrates the utility of transient current analysis for studying hDAT trafficking.
To increase the time resolution for detecting both changes in cell surface expression of hDAT and any modification in hDAT electrical activity, we voltage-stepped the membrane potential of the cell for 1 s to Ϫ80 mV every 1.5 s for a period of 45 s, during the decreasing phase of the AMPH-induced current (Fig. 7). We then analyzed the AMPH-induced currents by evaluating, over time, relatively small changes in current variance and charge movements comprising the steady-state and transient currents. Again, a strong correlation was found between the reduction in transient current and steady-state current over 30 consecutive voltage steps with no change in the dimension of the hDAT unitary current (Fig. 7).
We also used the ratio I/Q max to estimate the rate of activity of the individual transporters over time. At Ϫ160 mV, AMPH down-regulation of hDAT activity did not significantly alter the ratio I/Q max over the duration of the experiments. Therefore despite a decrease of the number of transporters on the cell surface (decrease in Q max ) the intrinsic activity of hDAT was unaltered. DAT has been shown to undergo constitutive recycling in PC12 cells (Loder and Melikian (23)), and it is unknown whether AMPH induces internalization, slows the return of constitutively internalized DAT to the plasma membrane, or both. Regardless, our data strongly suggest that AMPH downregulates cell surface expression and activity simultaneously with no change in the ability of the single transporter to conduct ions. Our results provide evidence with high time resolution (millisecond time scale) that hDAT is still active immediately prior to its cell surface redistribution in response to AMPH. Consequently, these data suggest that the regulation of hDAT trafficking may represent a cellular target for substance abuse therapies.