Na+ stimulates binding of dopamine to the dopamine transporter in cells but not in cell-free preparations

Authors


Address correspondence and reprint requests to Nianhang Chen, MD, Department of Biomedical and Therapeutic Sciences, University of Illinois College of Medicine, Box 1649, Peoria, IL 61656, USA. E-mail: nhc@uic.edu

Abstract

Although Na+ is crucial for the function of the dopamine (DA) transporter (DAT), its role in the substrate binding step has been questioned. To address this issue, we investigated the effect of Na+ on DA binding by measuring the potency of DA in inhibiting the binding of the cocaine analogue [3H]2β-carbomethoxy-3β-(4-fluorophenyl)tropane (CFT) in intact cells expressing DAT in their plasma membranes and in membranes isolated from these cells. In cells, Na+ substantially enhanced the potency of DA in inhibiting CFT binding. This effect of Na+ was independent of buffer compositions and substitutes (sucrose vs. NMDG), more pronounced at 4°C than 25°C, and correlated with its stimulatory effect on DA uptake Km. Removing extracellular Na+ had little effect on intracellular concentrations of Na+ and K+, or on membrane potential. These data suggest that extracellular Na+ most likely acts at the transporter level to enhance the binding of external DA during the transport cycle. In contrast, in cell-free membrane preparations the Na+ stimulation was abolished without impairment of the potency of DA in inhibiting CFT binding, regardless of whether sucrose was used to maintain the buffer osmolarity. The difference in Na+ dependence for DA to inhibit CFT binding between plasma membranes of intact cells and isolated membranes raises the possibility that intracellular ion environment, alone or in combination with other cellular factors, plays a critical role in determining DA–DAT interaction and the integration of Na+ modulation in this interaction.

Abbreviations used
CFT

2β-carbomethoxy-3β-(4-fluorophenyl)tropane

DA

dopamine

DAT

dopamine transporter

DiBAC2(3)

bis-(1,3-diethylthiobarbituric acid)trimethine oxonol

HEK

human embryonic kidney

HT

HEPES/Tris buffer

KRH

Krebs/Ringer/HEPES buffer

NMDG-Cl

N-methyl-d-glucamine chloride

PBFI

potassium-binding benzofuran isophthalate

SBFI

sodium-binding benzofuran isophthalate

The dopamine (DA) transporter (DAT) is a carrier protein located on the plasma membrane of DA nerve terminals. It is responsible for the re-uptake of DA from the synaptic cleft and extracellular space, and is the main mechanism for terminating DA neurotransmission in the central nervous system (Amara and Kuhar 1993; Gainetdinov and Caron 2002). The DAT is also a molecular target for drugs of abuse, such as cocaine and amphetamine, as well as for therapeutical drugs, such as tricyclic antidepressants (Amara and Sonders 1998; Chen and Reith 2002). Like other members of the Na+/Cl-dependent neurotransmitter transporter family, Na+ plays a critical role in the function of the DAT. Thus, the DAT requires external Na+ for inward transport (McElvain and Schenk 1992; Gu et al. 1994; Chen et al. 1999) and internal Na+ for reverse transport (Khoshbouei et al. 2003) of substrates. In addition to the Na+-influx coupled with substrate transport (Khoshbouei et al. 2003), the DAT also generates uncoupled currents that can be carried by Na+ (Sonders et al. 1997). It has also been suggested that Na+ binding to the DAT protects the transporter against the inhibition effect of cations present in extracellular fluid such as Ca2+, Mg2+ and K+ (Amejdki-Chab et al. 1992; Corera et al. 2000). However, a direct involvement of Na+ in substrate binding steps remains to be elucidated.

Although a direct test for the substrate binding at the DAT has recently been described using a fluorescent analogue of the neurotoxin MPP+ (Schwartz et al. 2003), until now there is no method to directly measure the binding of biogenic amines to monoamine transporters. In biochemical studies assessing DA binding, the measuring tool available so far is the DA-induced inhibition of binding of a radiolabeled blocker to the DAT. By measuring the potency of DA in inhibiting the binding of a non-transportable DAT blocker and cocaine analogue, [3H]2β-carbomethoxy-3β-(4-fluorophenyl)tropane (CFT), we have shown that substrate binding at the DAT in membrane preparations does not require Na+ and is not stimulated by Na+ (Chen et al. 2002; Li et al. 2002). However, we recognize that observations from experiments using membrane preparations may not necessarily represent what occurs at the plasma membrane of living cells. The transporter is an integral membrane protein with different parts exposed to different sides of the membrane. Thus, in cells, the extra- and intracellular portions of the transporter protein face different ionic environments with the cell interior containing regulatory signaling systems; in contrast, in isolated membranes all portions of the protein face the same medium devoid of signaling cascades. Further, in cells, the added Na+ is confined at the extracellular side of the plasma membrane and exogenous DA can be taken up, but in isolated membranes, added Na+ and DA gain free access to both sides. These differences may have profound influence on how the DAT interconverts between conformations and how Na+ and substrates reach the binding sites on the DAT. Indeed, in our preliminary studies, DAT mutations showed different effects on the interaction of the transporter with Na+ or DA depending on whether plasma membranes of intact cells or isolated membranes were studied. It therefore becomes important to understand the effect of Na+ on DA binding to the DAT at the cellular level. To address this issue, we used the [3H]CFT competition method, combined with functional and imaging approaches, to investigate the effect of Na+ on DA binding to intact HEK293 cells expressing DAT in their plasma membranes as well as in membranes isolated from these cells.

Materials and methods

Generation of cell lines stably expressing DAT

Stable pools of HEK293 cells expressing human DATs were obtained and maintained as described previously (Chen et al. 2001, 2002).

[3H]CFT binding to membrane preparations

Cell membranes were prepared as described previously (Chen et al. 2002) except that washes and re-suspension of membranes were done in a Na+-free buffer: HEPES-Tris buffer (HT) (10 mm/5 mm, pH 7.4) or the HT buffer containing 240 mm sucrose (HT/sucrose). For studies involving Na+ dependence, sodium isethionate, rather than NaCl, was used to avoid a potential effect of Cl on DA binding (Li and Reith 1999; Corera et al. 2000). Sodium isethionate showed no adverse effect on [3H]CFT binding (Chen et al. 2002). Membranes were incubated with 2 or 4 nm[3H]CFT (84.5 Ci/mmol, PerkinElmer Life Sciences, Boston, MA, USA) for 20 min at 25°C in a total volume of 200 μL. For saturation or competition analysis, varying concentrations of non-radioactive CFT (0.3–300 nm) or DA (0.3–1000 μm) were included in the assay. The binding reactions were terminated by filtration followed by five washes with ice-cold saline on 0.1% polythyleneimine pre-soaked Wallac B filtermats (Wallac, Gaithersburg, MD, USA) with a 96-well Brandel cell harvester (Gaithersburg, MD, USA). The radioactivity was counted in a 1405 Microbeta liquid scintillation counter (Wallac) after adding 10 mL of Betaplate Scint to each filter mat.

[3H]CFT binding to whole cells

Two buffer systems were used: one was HT/sucrose buffer, exactly the same as that for the membrane binding assays; the other was modified Krebs-Ringer-HEPES buffer (KRH), which contained 10 mm HEPES (adjusted to pH 7.4 with Tris), 130 mm NaCl unless otherwise stated, 1.2 mm KH2PO4, 1.2 mm MgSO4, 1 mm CaCl2, 3 mm KCl, 10 mm glucose, 0.1 mm tropolone and 1 mm ascorbic acid. For experiments determining Na+ dependence, Na+ concentration in the buffer was varied by adding 1–20 mm sodium isethionate to HT/sucrose, or by adding 1–20 mm NaCl to a Na+-free KRH supplemented with 130 mm N-methyl-d-glucamine-chloride (NMDG-Cl, KRH/NMDG), or by isotonically replacing NaCl in KRH with NMDG-Cl. Cell suspensions were prepared as described previously (Chen et al. 2001) except that the final wash and re-suspension of cells was done in the corresponding Na+-free assay buffer. Cells were incubated with 2–4 nm[3H]CFT for 2 h at 4°C or for 20 min at 25°C. The other procedures were the same as described for membrane binding assays.

[3H]DA uptake

Only KRH buffer was used and the Na+ concentration was varied by isotonically replacing NaCl with NMDG-Cl. Cell preparations were made as described for cell binding assays. Cells were incubated with 10 nm[3H]DA (60 Ci/mmol, PerkinElmer Life Sciences) for 5 min at 25°C in a total volume of 200 μL. For saturation analysis, various concentrations of DA (0.01–30 μm) were included in the assay. The other procedures were the same as described for membrane binding assays, except that the uptake reactions were terminated and washed on the harvester under lower pressure.

Ratiometric fluorescence measurements of intracellular Na+ and K+ concentrations

Serum-starved cells grown on the glass bottom of a Petri dish were loaded with the cell-permeant acetoxymethyl ester (AM) forms of SBFI (10 μm) and PBFI (5 μm) (Molecular Probes, Eugene, OR, USA) in the presence of 0.02% Pluronic F-127 (Molecular Probes) and 1% BSA for 2 h at 25°C. The cells were then washed with 1% BSA-containing KRH twice and incubated for 30 min in regular KRH to allow de-esterification of the fluorophores. Afterwards, cells were incubated in KRH containing 130 mm NMDG-Cl (Na+-free) or NaCl for 20 min. Measurements of [Na+]i and [K+]i were performed using a dual-excitation fluorescence imaging system (Intracellular Imaging, Cincinnati, OH, USA) in conjunction with a Nikon TMS-F invert microscope. SBFI- or PBFI-loaded cells were alternately excited via a Nikon Fluor 20 Ph3DL objective at 340 and 380 nm by a computer-controlled filter changer and the fluorescence emissions from individual cells were monitored at 510 nm by a COHU High-performance CCD camera. Sequential image pairs were collected every 10 s for 10 min followed by in situ calibration. Calibration of the excitation ratio in terms of the ion concentration was accomplished by equilibrating the dye-loaded cells with a KRH buffer containing various Na+ or K+ concentrations (0, 3, 10, 30, 70 and 140 mm), 10 μm gramicidin, and 100 μm ouabain. The Na+ calibration solutions were made by mixing NaCl and KCl to a total concentration of 140 mm, and the K+ calibration solution, by mixing KCl and LiCl. Experiments were repeated on three different coverslips, each allowing collection data from 10 individual cells or more. Experimental F340/F380 values were converted to intracellular concentrations utilizing the equation [ion]i = Km*(R − R0)/(Rmax − R), where R is the experimental F340/F380 ratio, R0 is the fluorescence ratio in the absence of the ion, Rmax is the fluorescence ratio of the ion-saturated dye, and Km is the Michaelis-Menten constant that reflects the affinity of the dye for that ion.

Fluorescent imaging of membrane potential changes

The lipophilic anionic fluorescent dye DiSBAC2(3) was used to detect changes in membrane potential. This dye is reported to accumulate in the cytoplasm of depolarized cells via a Nernst equilibrium-dependent uptake (Dall'Asta et al. 1997). Serum-starved cells grown on the glass coverslip were pre-incubated with 1 μm DiSBAC2(3) in KRH at 25°C for 1 h, and then subjected to KRH buffer containing 130 mm NMDG-Cl or NaCl for 30 min in the presence of 1 μm DiSBAC2(3). DiSBAC2(3) fluorescence was viewed with an Olympus BX61 fluorescence microscope equipped with a neutral density filter U-25ND25, excitation filter BP 470–490, and barrier filter BA510-550. Images were captured with a CCD camera operated with Spot software (Olympus, Melville, NY, USA).

Data analysis

Cells transfected with vector alone were tested in parallel and used to estimate non-specific binding or uptake. Non-linear regression fitting of data was performed using LIGAND (Biosoft, Cambridge, UK) and ORIGIN (OriginLab Co., Northampton, MA, USA) to estimate Kd, Bmax, Km, Vmax, EC50 and IC50 values. Apparent Ki values were calculated by the Cheng-Prusoff equation.

Results

Ambient Na+ has no effect on inhibition of CFT binding by DA in membrane preparations

The effect of Na+ on DA Ki for inhibiting [3H]CFT binding was initially tested in membrane preparations where both sides of the membrane face the same ion environment. Both CFT binding and inhibition of this binding by DA could be readily measured in the absence of Na+ with HT or HT/sucrose buffers. However, only the Kd of CFT binding was sensitive to Na+. At 20 mm, a concentration producing a plateau in stimulating CFT binding, Na+ reduced the CFT Kd by 10-fold, but had no effect at all on the DA Ki, regardless of the presence of sucrose (Table 1). Further raising Na+ to 150 mm or using KRH buffer also did not modify the Ki for DA (data not shown). Because the inhibitory effect of other Na+ substitutes made CFT binding undetectable in a Na+-free buffer, the use of other substitutes for Na+ was not attempted in these experiments.

Table 1.  Effect of Na+ on affinity of [3H]CFT binding and potency of DA in inhibiting [3H]CFT binding
Parameters & preparationsTemperatureBuffer/substitute[Na+], mmFold decrease
01201/0 mm20/0 mm
  1. Cells or membrane preparations were incubated with 2 or 4 nm[3H]CFT, various concentrations of unlabeled CFT or DA for 20 min at 25°C or 2 h at 4°C in HT or KRH buffer containing substitutes for Na+ as indicated (240 mm sucrose or 130 mm NMDG-Cl). The indicated Na+ concentrations were achieved by adding NaCl (to KRH) or sodium isethionate (to HT) directly. Data are means ± SE for more than four experiments, each performed in triplicate. *p < 0.05 versus the corresponding group in the absence of Na+ (Dunnett's-test).

CFT Kd, nm
Membranes25°CHT95.3 ± 9.54 8.54 ± 0.56* 11.2 ± 1.33
Membranes25°CHT/sucrose88.8 ± 10.244.8 ± 4.09*7.48 ± 0.80*2.03 ± 0.1711.9 ± 1.85
Cells25°CKRH/NMDG39.0 ± 4.4833.8 ± 3.4914.6 ± 2.60*1.19 ± 0.112.90 ± 0.42
Cells25°CHT/sucrose19.5 ± 2.068.69 ± 1.74*4.15 ± 1.38*3.24 ± 0.816.56 ± 0.94
Cells 4°CKRH/NMDG64.4 ± 3.6664.1 ± 10.425.5 ± 3.41*1.10 ± 0.192.71 ± 0.45
Cells 4°CHT/sucrose58.9 ± 7.3820.6 ± 2.66*5.18 ± 0.78*3.16 ± 0.4812.5 ± 2.44
DA Ki, μm
Membranes25°CHT5.01 ± 0.99 5.36 ± 0.76 1.01 ± 0.11
Membranes25°CHT/sucrose4.42 ± 0.324.62 ± 0.955.48 ± 0.881.08 ± 0.200.87 ± 0.13
Cells25°CKRH/NMDG41.8 ± 8.7014.8 ± 1.19*9.20 ± 0.70*2.90 ± 0.254.62 ± 0.33
Cells25°CHT/sucrose8.93 ± 0.384.33 ± 0.74*1.92 ± 0.22*2.51 ± 0.325.90 ± 0.13
Cells 4°CKRH/NMDG34.2 ± 1.0219.4 ± 2.413.44 ± 0.42*1.86 ± 0.2410.4 ± 1.25
Cells 4°CHT/sucrose31.2 ± 8.404.08 ± 1.21*1.78 ± 0.52*11.3 ± 2.6423.2 ± 6.77

External Na+ enhances inhibition of CFT binding by DA in cells

The effect of Na+ on DA Ki was then tested in cells where the surface membrane exhibits restricted sidedness and ion accessibility. We used two buffer systems. The simple HT/sucrose contained no cations and allowed us to compare the results from cells with those from membrane preparations. The KRH/NMDG contained physiological concentrations of other cations (Ca2+, Mg2+ and K+), and also allowed us, in cell binding assays, to compare the results with sucrose as a Na+ substitute with those with NMDG as the substitute. We reasoned that if Na+ directly enhances the DA–DAT interaction, a reduction in the DA Ki in inhibiting CFT binding would be observed irrespective of buffer systems and substitutes used. To facilitate interpretation of the results, sucrose was fixed at 240 mm or NMDG-Cl at 130 mm, and Na+ concentration was varied between 0 and 20 mm by adding sodium isethionate (to HT/sucrose) or NaCl (to KRH/NMDG) directly. Under all conditions tested, CFT binding reached equilibrium within the incubation time used, and the binding was best fit by a one-site model (Fig. 1). Increasing Na+ concentration reduced Kd (Fig. 1 and Table 1) without substantial effect on Bmax (Table 2).

Figure 1.

Scatchard plots of [3H]CFT binding to cells expressing DAT in the presence of various concentrations of Na+. (a) In KRH/NMDG at 4°C; (b) in HT/Sucrose at 4°C; (c) in KRH/NMDG at 25°C, (d) in HT/Sucrose at 25°C. The indicated Na+ concentrations were achieved by adding NaCl (to KRH/NMDG) or sodium isethionate (to HT/Sucrose). The solid straight line represents the best fit chosen by ligand program. Each panel shows a representative experiment performed in triplicate with three levels of [Na+] on the same batch of cells. Each experiment was repeated more than three times.

Table 2.  Effect of Na+ on maximal capacity for [3H]CFT binding to cells expressing DAT
Temperature[Na+]
(mm)
KRH/NMDG
(pmol/mg)
HT/sucrose
(pmol/mg)
  1. Cells were incubated with 2 or 4 nm[3H]CFT, various concentrations of unlabeled CFT for 20 min at 25°C or 2 h at 4°C in HT or KRH buffer containing substitutes for Na+ as indicated (240 mm sucrose or 130 mm NMDG-Cl). The indicated Na+ concentrations were achieved by adding NaCl (to KRH/NMDG) or sodium isethionate (to HT/Sucrose). Data are means ± SE for more than four experiments, each performed in triplicate.

25°C 02.96 ± 0.415.38 ± 0.25
 13.34 ± 0.293.74 ± 0.44
203.73 ± 0.353.41 ± 0.24
4°C 03.56 ± 0.213.89 ± 0.58
 13.36 ± 0.513.48 ± 0.38
203.72 ± 0.433.82 ± 0.30

Similar to what was observed in membrane preparations, DA inhibited CFT binding to cells even in the absence of Na+, but the potency was significantly lower (Table 1). Different from what was observed in membrane preparations, in cells, increasing Na+ concentration considerably reduced the DA Ki for inhibiting CFT binding. Noticeably, in most cases, this effect became statistically significant at the concentration as low as 1 mm. Moreover, the DA Ki-reducing effect of Na+ was consistently observed regardless of substitutes (sucrose or NMDG-Cl) and ion components in the buffer (simple HT or multiple cation-containing KRH).

To assess at which step Na+ enhances the inhibition of CFT binding by DA, we examined the effect of temperature on the Na+ dependence of DA Ki. Remarkably, at 4°C where a substrate is not substantially transported, in both buffer systems, addition of Na+ up to 20 mm produced a greater reduction in the DA Ki for inhibiting [3H]CFT binding than it did at 25°C (Table 1). It was also noted that at 4°C the fractional reduction in DA Ki induced by 20 mm Na+ was substantially greater than that in CFT Kd (10–20-fold vs. 2–10-fold).

Substitute: sucrose versus NMDG

In membrane preparations, the addition of 240 mm sucrose did not alter the values for CFT Kd or DA Ki in the absence of Na; and it also did not alter the effect of Na+ on these values (Table 1), suggesting that sucrose does not interact with the DAT or Na+ under this condition. In cells, however, there were differences between substitutions with sucrose and NMDG. Firstly, in the absence of Na+, although the values for CFT Kd or DA Ki were virtually the same with sucrose or NMDG substitutions at 4°C, the values became smaller in sucrose than in NMDG at 25°C (Table 1). Secondly, at both 4°C and 25°C, CFT Kd was reduced by Na+ more drastically in the presence of sucrose than in the presence of NMDG (Table 1). At 4°C, the Na+-induced fractional reduction in DA Ki was also much greater in the presence of sucrose than in the presence of NMDG (Table 1). Thirdly, at 1 and 20 mm Na+, the values for CFT Kd and DA Ki were significantly lower in the presence of sucrose than in the presence of NMDG regardless of assay temperatures (Table 1). To determine whether the substitutes interact with Na+, we examined [3H]CFT binding as a function of Na+ in the presence of 240 mm sucrose or with Na+ isotonically replaced by NMDG in the same TH buffer. As shown in Fig. 2(a), with NMDG as the Na+ substitute, CFT binding approached its plateau at 20 mm Na+ plus 110 mm NMDG. This suggests that NMDG, at least up to 110 mm, had little inhibitory effect. Compared with NMDG substitution, the Na+ EC50 was smaller and the percentage maximal stimulation was higher in the presence of 240 mm sucrose. In support for a stimulatory interaction between sucrose and Na+, the CFT binding level (Fig. 2b) was significantly higher and the DA Ki value (Fig. 2c) was markedly smaller at 20 mm Na+ plus 240 mm sucrose than at 130 mm Na+ alone.

Figure 2.

Interactions of sucrose with Na+. (a) [3H]CFT binding to cells as a function of Na+ concentration with NMDG+ or sucrose as a Na+ substitute. The EC50 is the concentration for Na+ to produce half-maximal stimulation of [3H]CFT binding. (b) [3H]CFT binding at 130 mm Na+ or at 20 mm Na+ plus 240 mm sucrose. (c) DA Ki for inhibiting the CFT binding at 130 mm Na+ or at 20 mm Na+ plus 240 mm sucrose. The indicated Na+ concentrations were achieved by isotonically replacing NaCl with NMDG-Cl in HT buffer or directly adding sodium isethionate to HT/Sucrose buffer. Data are means ±SE for more than four experiments, each performed in triplicate. *p < 0.05 versus Na+ 130 mm group (t-test).

Na+ dependence of DA Ki in inhibiting CFT binding and of DA uptake

Because of its relative inert feature as well as the advantage of maintaining a constant concentration of Cl ions, NMDG-Cl was used to replace NaCl in a wider range in further experiments with cell preparations. In binding assays at 25 or 4°C, an increase in Na+ concentration from 0 to 130 mm caused a 7- or 30-fold reduction in DA Ki: from 39 ± 4.48 to 5.41 ± 0.55 μm at 25°C, and from 64.4 ± 3.66 to 2.11 ± 0.12 μm at 4°C. To address the functional relevance of the Na+ dependence of DA Ki in inhibiting [3H]CFT binding, we compared the Na+-induced fractional change in DA Ki with that in Km for DA uptake in a Na+ concentration range of 5–130 mm. It is apparent from Fig. 3 that there was a close correlation between Na+-induced changes in Ki and Km, especially at identical temperature (25°C for both assays). This indicates that the Na+ dependence of DA Ki in inhibiting CFT binding is highly relevant to events allowing Na+-dependent dopamine binding during transport. The Vmax for DA uptake also increased upon raising the Na+ concentration to 130 mm, probably due to the fact that Na+ is co-transported with the substrate.

Figure 3.

Na+ dependence of [3H]DA uptake and potency of DA in inhibiting [3H]CFT binding. (a) 25°C for [3H]DA uptake and 4°C for [3H]CFT binding at 10, 40, 80 and 130 mm Na+. (b) 25°C for both [3H]DA uptake and [3H]CFT binding at 5, 20, 60 and 130 mm Na+. All assays were done in KRH buffer with NaCl isotonically replaced by NMDG-Cl. The results are expressed as a percentage of the value at the lowest Na+ tested. Data are means ± SE for six experiments, each performed in triplicate. *p < 0.05 between the value at the lowest concentration of Na+ and the values at higher concentrations of Na+ (Dunnett's-test).

Intracellular ion concentration and membrane potential upon varying extracellular Na+ concentration

In in situ calibrations, the Km of SBFI for Na+ was 17.2 ± 1.2 mm and the Km of PBFI for K+ was 5.88 ± 0.24 mm. The intracellular concentration of Na+ in single cells ranged between 0.5 and 5 mm with an average level of ∼3 mm under control condition, and was not markedly changed by removing extracellular Na+ (Table 3). There was also no detectable change in the intracellular K+ concentration upon removing the extracellular Na+ concentration (Table 3). Based on the known intracellular K+ concentration and the buffer K+ concentration, the resting membrane potential was calculated with the Nernst equation. The membrane potential was similar at 0 and 130 mm Na+ (Table 3). However, due to the high affinity of PBFI for K+, when intracellular [K+] exceeded 70 mm, PBFI was practically saturated with K+, and modest changes in intracellular [K+] might not be detected. Therefore, the lipophilic anionic fluorescent dye DiSBAC2(3) was also used to visualize membrane potential changes. This dye can accumulate in the cytoplasm of depolarized cells, resulting in enhanced intensity of the DiSBAC2(3) fluorescence (Dall'Asta et al. 1997). In support for unchanged intracellular [K+] and membrane potential, the DiSBAC2(3) fluorescence in single cells showed similar low intensity between 0 and 130 mm extracellular Na+(Figs 4a and b). In contrast, depolarization with 30 mm K+ resulted in enhanced cellular DiSBAC2(3) fluorescence (Fig. 4c). Thus, observed changes in DA Ki upon varying [Na+] are not due to changes in intracellular concentrations of Na+ and K+ or membrane potential.

Table 3.  Intracellular ion concentrations and membrane potential in single cells at different extracellular Na+ concentrations
[Na+]o, mm[K+]o, mm[Na+]i, mm[K+]i, mmEm, mV
  1. SBFI or PBFI loaded cells were incubated at 25°C in KRH containing 130 mm NMDG-Cl or NaCl for 20 min. Afterwards, the fluorescence ratios were recorded for 10 min and the recordings during the last 5 min were averaged and converted to intracellular concentrations as detailed in Materials and methods. Experiments were repeated on three different coverslips, each allowing collecting data from 10 individual cells or more. Membrane potential was calculated as Em = 59 log [K+]o/[K+]i. The number in brackets represents number of cells tested. Data are means ± SE.

032.21 ± 0.17 (32)144 ± 12 (32)− 99
13032.90 ± 0.32 (36)139 ± 7 (37)− 98
Figure 4.

Cellular DiSBAC2(3) fluorescence emission at different extracellular Na+ concentrations. (a) 130 mm NMDG; (b) 130 mm Na+ (c) 30 mm K+ (plus 100 mm Na+). Cells were pre-loaded with 1 μm DiSBAC2(3) in KRH at 25°C for 1 h. The medium was then removed and cells were incubated in KRH containing the indicated ion concentrations in the presence of 1 μm DiSBAC2(3) for 30 min. Images were taken with a 4-s exposure. The experiments were repeated three times with comparable results.

Discussion

To validate a direct regulation by external Na+ of DA binding to DAT on the plasma membrane of intact cells, one needs to establish that the effect is solely due to the presence of external Na+, independent of inward translocation cycles coupled with external Na+, and unrelated to alterations in cellular millieu associated with extracellular Na+. These issues have been addressed by the present study.

For investigation of Na+ dependence, the ideal substitute for Na+ to maintain the buffer osmolarity should be inert with regard to the protein studied, but mimic Na+ in other aspects of its biochemical and biophysical properties. Unfortunately, almost all of the commonly used substitutes (Li+, Tris, choline, NMDG and sucrose) show a more or less pronounced effect on DAT activity, depending on the preparation carrying the DAT. In general, potent inhibition of the transporter by cationic substitutes is observed in membrane preparations (Amejdki-Chab et al. 1992; Reith and Coffey 1993; Chen et al. 2002; Li et al. 2002) but not in cells (Wu et al. 1997). Though other possibilities exist, one interpretation for this difference is that cationic substitutes interact with certain DAT domains that are accessible in isolated membranes but inaccessible from the extracellular side of the plasma membrane of cells. Interactions between Na+ and some cationic substitutes at the DAT level have also been observed in membrane preparations (Chen et al. 2002) but remain unclear in cells. Thus, when stimulation is found upon increasing [Na+] while reducing the Na+ substitute, one needs to consider the possibilities of (i) Na+ stimulation, (ii) simple removal of the inhibitory substitute, and (iii) interaction between Na+ and the substitute. Two strategies were employed to distinguish these possibilities. Firstly, we added modest concentrations of Na+ to a constant high concentration of the substitute, so that any observed changes can only be attributed to the addition of Na+. Secondly, we tested the effect of Na+ in the presence of two different substitutes, sucrose and NMDG, to rule out potential interactions between Na+ and substitutes. Unlike commonly used cationic substitutes that reduced CFT binding to an undetectable level in membrane preparations in a Na+-free buffer (Chen et al. 2002), sucrose did not have direct effect on DAT in membrane preparations, which allowed us to compare the Na+ dependence between membrane preparations and cells under the same conditions. However, in cells, sucrose did not appear inert as shown by its ability to stimulate binding in a Na+-free buffer in a temperature-sensitive manner and by its co-operative interaction with Na+. Thus, in experiments involving cells, we also examined the effect of Na+ using NMDG as the substitute. Although it inhibited CFT binding to membrane preparations (Chen et al. 2002), NMDG was relatively inactive in cells. In Na+-free buffer, the presence of NMDG allowed easy detection of DA–DAT interaction in a temperature-insensitive manner. Under this condition, the values for DA Ki obtained with 130 mm NMDG were similar to those with 240 mm sucrose at 4°C and did not change when the temperature was raised to 25°C. Thus, in comparison to sucrose, NMDG is neither inhibitory nor stimulatory in cells. Moreover, NMDG substitution for Na+ allowed CFT binding to saturate in the submillimolar range of Na+ concentrations (Fig. 2a), reminiscent of the situation in membrane preparations with a buffer free of inhibitory Na+ substitutes (Chen et al. 2002). Though sucrose and NMDG have quite different properties, with either substitute fixed at a high concentration, we found that the addition of Na+ always reduced the DA Ki, regardless of buffer components and assay temperatures (Table 1). Therefore, we conclude that Na+ itself, rather than an interaction between Na+ and the substitute, is responsible for the enhanced ability of DA to inhibit CFT binding to DAT-expressing cells.

The Na+-induced reduction in DA Ki for inhibiting CFT binding to DAT-expressing cells could be due to a stimulatory effect of Na+ on DA binding. It could also be due to an effect of Na+ on transport if conformational changes associated with substrate translocation were unfavorable to CFT binding to DAT. However, there is no substantial substrate transport at 4°C (Chen et al. 2000). The observation that the effect of Na+ on DA Ki was more pronounced at 4°C is consistent with the assertion that Na+ enhances the inhibition of CFT binding by DA mainly via enhancing DA interactions at binding-associated steps. We also show that Na+-induced changes in DA Ki for inhibiting CFT binding were in parallel to those in Km for DA uptake. Such a close correlation in the Na+ profile strongly supports the interpretation that the Na+-dependent inhibition of CFT binding by DA reflects the Na+-dependent DA binding linked to the substrate transport.

Unlike the situation with isolated membranes, DA–DAT interactions at the plasma membrane of intact cells may be under the influence of transmembrane ion gradients and membrane potential. In neuronal cells, a change in the transmembrane Na+ gradient by varying buffer Na+ concentrations might alter the neuronal excitability and membrane potential (Hille 1992). Such a possibility appears remote in the present non-excitable HEK293 cells. Indeed, from the lack of change in cytoplasmic appearance of the fluorescent and potential-sensitive probe DiSBA2(3), it may be inferred that alterations in extracellular [Na+] does not change the membrane potential. In support, the K+ diffusion potential, as calculated from Nernst equation, was the same in the absence and presence of Na+. Further, the near normal level of intracellular [Na+] upon removal of extracellular Na+ indicates that external Na+ does not exert its stimulatory effect by altering the intracellular level of Na+. Another concern is that intracellular signal transduction systems are operative in cells but lost in membrane preparations. Compelling evidence indicates that the availability of DAT at the cell surface is tightly regulated by receptor- or PKC-mediated transporter internalization (Blakely and Bauman 2000; Zahniser and Doolen 2001; Robinson 2002). It could be contemplated that changes in external Na+ levels might modify these regulation processes. However, the temperature independence for Na+ to enhance DA–DAT interactions argues against an involvement of intracellular enzyme-mediated regulatory mechanisms. We also found that the distribution pattern of transporter immunofluorescence on the cells surface was virtually the same in the absence or presence of extracellular Na+ (Chen et al. work in progress). Therefore, it is unlikely that external Na+ regulates DA–DAT interaction by modifying other associated cellular processes.

Taken together, our data point to a direct action of extracellular Na+ at the transporter level. Previously, Sonders et al. (1997) demonstrated that constitutive DAT-mediated leak currents in oocytes were blocked by DA in the absence of Na+, suggesting that Na+ is not absolutely required for DA binding to DAT. Our findings are consistent with this report in that DA still inhibited CFT binding in the absence of Na+. However, an increase in Na+ from 0 to 130 mm resulted in an approximately 7- to 20-fold higher potency for DA, underscoring the physiological significance of Na+ regulation. In the alternating model proposed for monoamine transporters (Rudnick 2002), the transporter protein is thought to contain a central binding site for the substrate, which is exposed only to one side of the membrane at a time via conformational transition. It is possible that the DAT spontaneously interconverts between multiple conformations, and external DA interacts only with (a) certain conformation(s). Extracellular Na+ could act at the DAT to facilitate conversion to the conformation favoring external DA binding. In this view, the DA Ki is a function of the fraction of DATs residing in the state favoring external DA, with the Ki measurement representing the time average of the different states occurring during ongoing interconversions.

The present study reveals that DA–DAT interaction at the plasma membrane of cells has an Na+-dependence not shared by isolated membranes. This difference suggests that DA binding to isolated membranes differs from that to plasma membranes of intact cells. Intriguingly, in the cell-free DAT preparation Na+ dependence was abolished without impairment of the ability of DA to inhibit CFT binding. As judged from Na+-dependent CFT binding, Na+ can bind to the DAT and can still induce conformational changes in DAT in membrane preparations. Thus, the potent and yet Na+-insensitive DA Ki in membrane preparations suggests that, in this case, DA can readily bind different states of DAT with the same affinity. The occurrence of this would be more likely when the various DAT states have a common binding site that DA gains easy access to. In membranes with both the extra- and intracellular portions of DAT exposed to the same ionic environment, the various transporter states could differ from those on plasma membrane of intact cells. The substrate site might therefore be exposed independent of Na+-associated conformational changes in membrane preparations. Alternatively, the exposure of the intracellular portion of DAT to the bathing medium might provide an opportunity for ambient DA to reach the binding site predominantly from the intracellular side of the membrane, and this process may be Na+ independent. More work is needed to delineate the mechanisms that determine the difference between intact cells and membranes in the Na+ dependence of substrate binding.

Acknowledgements

The study was supported by National Institute on Drug Abuse Grants DA08379 and 13261.

Ancillary