SEARCH

SEARCH BY CITATION

Keywords:

  • binding;
  • cation;
  • cocaine analog;
  • dopamine transporter;
  • protein conformation

Abstract

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements

In membrane preparations, CFT, a phenyltropane cocaine analog, and dopamine (DA) interact with the recombinant human dopamine transporter (hDAT) in Na+-free medium. Na+ markedly increased the transporter's affinity for CFT, but had little or no effect on DA potency for inhibiting CFT binding. Raising [Na+] from 20 to 155 mm reduced Li+-induced increase in DA Ki, but not CFT Kd. The presence of 155 mm Na+ enhanced the tolerance to low pH of CFT Kd but not DA Ki. Leucine substitution for tryptophan 84 (W84L) in transmembrane domain (TM) 1 or asparagine substitution for aspartate 313 (D313N) in TM 6 did not or only modestly enhance the affinity of Na+-independent CFT binding, and retained the near normal ability of DA, Li+, K+, or H+ to inhibit this binding. However, the mutations significantly enhanced the Na+ stimulation of CFT binding as well as the Na+ antagonism against Li+ and H+ inhibition of CFT binding. In contrast, the mutations neither changed the Na+-insensitive feature of DA Ki nor enhanced the Na+ protection of DA Ki against Li+'s inhibitory effect, though they caused Na+ protection of DA Ki against H+'s inhibitory action. These results are consistent with the existence of binding conformations for DA that are distinguishable from those for CFT, and with a differential association of cation interactions with DA and CFT binding. The mutations likely alter Na+-bound state(s) of hDAT, preferentially strengthening the positive allosteric coupling between Na+ and CFT binding, and reducing the impact of Li+ or H+ on the CFT binding.

Abbreviations used:
CFT

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

DA

dopamine

GABA

γ-aminobutyric acid

hDAT

human dopamine transporter

TM

transmembrane domain

Na+–ISE

sodium isethionate

NMDG

N-methyl-d-glucamine.

The involvement of monovalent alkali cations and protons in interactions of the human dopamine (DA) transporter (hDAT) with substrates and cocaine analogs has been documented extensively. Like other members of the Na+/Cl-dependent neurotransmitter transporter family, the hDAT not only requires Na+ for uphill transport of substrates, but also generates a transport-associated current carried by Na+ as well as a leak current that can be carried by multiple ions such as K+, Li+, H+, and Na+ (Sonders et al. 1997). Despite their similarities in carrying the leak current that can be blocked by both substrates and inhibitors of the transporter, Na+ mainly promotes cocaine analog binding and substrate transport, but K+, Li+, and H+ inhibit these processes (Shank et al. 1987; Amejdki-Chab et al. 1992; Xu and Reith 1996; Chen et al. 1997, 1999; Berfield et al. 1999; Li and Reith 1999). Interactions of Na+ with other cations at the transporter level have been observed in several members of the Na+/Cl-dependent transporter family. Thus, Na+ prevents K+ inhibition of cocaine analog and DA binding in rat and human DATs (Chen et al. 1997; Li and Reith 1999, 2000), and antagonizes the effect of Li+ on chemical inactivation and current in serotonin transporters (Ni et al. 2001). Additionally, in serotonin and GABA transporters, the mutation-induced change in Na+ sensitivity can be accompanied by changes in pH sensitivity (Barker et al. 1999; Forlani et al. 2001). It has been demonstrated that the Na+ binding at the transporter protein can bring about conformational changes (Mager et al. 1996; Loo et al. 1998; Maehrel et al. 1998; Li et al. 2000). Recently, antagonism between Na+ and Li+ at the serotonin transporter has been suggested to result from different conformational changes induced by the two cations (Petersen and DeFelice 1999; Kamdar et al. 2001; Ni et al. 2001).

Kinetic analysis of DA transport (McElvain and Schenk 1992; Chen et al. 1999; Earles and Schenk 1999) and mathematical modeling of cocaine analog and substrate binding (Chen et al. 1997; Li and Reith 1999, 2000) have provided evidence implicating different Na+ requirements between substrates and cocaine analogs. In our mutagenesis studies on negatively charged residues and dipole tryptophan residues that are highly conserved in the Na+/Cl-dependent neurotransmitter transporter family, mutations often differentially modified cation-sensitivity of cocaine analog binding and DA uptake (Chen et al. 2001a). Moreover, some of these mutations also shifted the relative selectivity of the transporter towards either cocaine analogs or DA. We speculated that the binding of cocaine analogs and DA involve non-identical conformations differentially coupled to cation-binding sites. Thus, a shift in ligand selectivity by a mutation could be caused by alterations in cationic interactions that are critical in directing the binding preference of the hDAT. The present study was initiated to address this hypothesis by assessing the effect of individual cations Na+, K+, Li+, and H+, as well as the impact of Na+ interactions with Li+ or H+, on the binding of a phenyltropane analog of cocaine, 2β-carbomethoxy-3β-(4fluorophenyl)tropane (CFT), and DA. DA binding was assessed indirectly through inhibition of [3H]CFT binding. We sought to distinguish cationic interactions with CFT binding from those with DA binding at wild-type as well as mutated hDATs. Two mutated hDATs were used in this study: leucine substitution for tryptophan 84 (W84L) in transmembrane domain (TM) 1 or asparagine substitution for aspartate 313 (D313N) in TM 6. Both residues are highly conserved among the member of Na+/Cl-dependent neurotransmitter transporter family. Their sequence conservation could reflect their involvement in features common for the Na+/Cl-dependent transporters, such as ion regulation. Moreover, both mutations resulted in not only changes in cationic sensitivity (Chen et al. 2001a) but also a shift in the relative selectivity towards cocaine analogs (Chen et al. 2001b). Knowledge regarding unique cationic susceptibility for cocaine analog and substrate binding could open the possibility to develop anticocaine medications aimed at the cation-modulating site on the hDAT.

Materials and methods

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements

Materials

The pCIN4 vector and the wild-type pCIN4-hDAT construct were generous gifts from Dr Jonathan A. Javitch of Columbia University. [3H]CFT (84.5 Ci/mmol) were purchased from NEN Life Sciences Products (Boston, MA, USA). Other chemicals were provided by commercial sources.

Generation of cell lines stably expressing wild-type and mutant hDATs

Site-directed mutagenesis of the hDAT was performed using the QuickChange Mutagenesis Kit (Stratagene, La Jolla, CA, USA). The mutants were screened by restriction mapping, and the mutation was confirmed in both directions by dye terminator cycle sequencing the complete coding region (Research Resources Center, University of Illinois).

Human embryonic kidney cells (HEK 293, ATCC CRL 1573) were transfected with the wild-type or mutant pCIN4-hDAT, or pCIN4 vector alone by using LipofectAMINE (Life Technologies, Inc., Gaithersburg, MD, USA) as described previously (Chen et al. 2001b). Transfected cells were selected in the growth medium containing 800 µg/mL geneticin, and afterwards maintained in the growth medium containing 200 µg/mL geneticin.

[3H]CFT binding to membrane preparations

After being lysed in 2 mm HEPES and 1 mm EDTA for 10 min at 4°C, transfected cells were scraped from plates and centrifuged at 20 000 g for 20 min at 4°C. The membrane pellet was homogenized (Brinkmann Polytron, setting 6 for 15 s) in and washed with ice-cold HEPES/Tris buffer (10 mm/5 mm, pH 7.4). The washed membranes were resuspended in appropriate buffers and used for assays. For studies involving cationic interactions, cations in chloride salts or sodium isethionate (Na+–ISE) were added to the HEPES/Tris buffer (pH 7.4, or as indicated, adjusted with Tris, NaOH, or HCl). All experiments on DA with Na+ were done with Na+–ISE except for those presented in Fig. 8. Cl contributed by NaCl has been shown to stimulate DA binding (see Li and Reith 1999; Corera et al. 2000), though our preliminary experiments did not show a significant effect of Cl on the DA binding under conditions of the experiments covered in this study. The effect of Na+–ISE on [3H]CFT binding was similar to NaCl (see Results).

image

Figure 8. Effect of antagonism between Na+ and H+ on interactions of CFT and DA with [3H]CFT labeled sites on wild-type and mutant hDATs. (a,d) Wild-type (WT); (b,e) W84L; (c,f) D313N. Cell membranes were incubated with 4 nm[3H]CFT for 15 min at 21°C in HEPES–NaOH buffer containing 5 or 155 mm NaCl at pH 6.4 and 7.4. Values represent means ± SE for 4–8 experiments performed in triplicate. *p < 0.05 versus pH 7.4 group at corresponding [Na+] (paired t-test after log transformation of data).

Download figure to PowerPoint

Membranes were incubated with 4–16 nm[3H]CFT for 15 min at 21°C in a total volume of 200 µL. For saturation or competition analysis, appropriate concentrations of non-radioactive CFT (0.3–300 nm) or DA (0.3–300 µm) were included in the assay. The binding reactions were terminated by filtration followed by five washes with ice-cold phosphate-buffered saline (PBS) on 0.1% polythyleneimine pre-soaked glass fiber filters using a 96-well Tomtec cell harvester (Wallac, Gaithersburg, MD, USA) or a 96-well Brandel cell harvester (Gaithersburg, MD, USA). The radioactivity in 5 mL/filtermat Betaplate Scint (Wallac) was counted in a 1405 Microbeta liquid scintillation counter (Wallac).

Data analysis

Membranes from cells transfected with pCIN4 vectors alone were tested in parallel and used to estimate non-specific binding. Nonlinear regression fitting of data was performed using LIGAND (Biosoft, Cambridge, UK) and ORIGIN (Microcal Software, Inc., Northampton, MA, USA) to estimate Kd, Bmax, EC50, and IC50 values. Apparent Ki values were calculated by the Cheng–Prusoff equation (Cheng and Prusoff 1973).

Results

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements

Effect of Na+ on [3H]CFT binding

With or without substitutes

The effect of Na+ was examined by directly adding NaCl or Na+–ISE to the HEPES–Tris buffer (Fig. 1a). [3H]CFT binding was enhanced by Na+ in an identical manner with either Cl or ISE as the counterion. This binding reached plateau between 20 and 40 mm Na+ (Fig. 1a). In contrast, when NaCl was isotonically replaced by NMDG–Cl or LiCl, the CFT binding showed significantly less enhancement by Na+ and did not reach plateau at 155 mm Na+ (Fig. 1b). Because of the apparent inhibitory effect of NMDG on the CFT binding under current conditions, the attempt to use NMDG as a Na+ substitute was abandoned.

image

Figure 1. [3H]CFT binding as a function of Na+ concentrations with or without Na+ substitutes. (a) No substitutes; (b) NMDG or Li+ as substitutes. Membranes of cells expressing wild-type hDATs were prepared in Na+-free HEPES–Tris buffer, and incubated with 4 nm[3H]CFT for 15 min at 21°C in the presence of various concentrations of Na+. Na+ concentrations were obtained by directly adding NaCl or Na+–ISE (a), or by isotonically replacing NaCl with NMDG-Cl or LiCl (b). Values represent means ± SEM of 2–4 experiments performed in triplicate.

Download figure to PowerPoint

CFT binding as a function of [Na+]

At a range of concentrations up to 10–40 mm (depending on the hDATs tested), Na+ stimulated [3H]CFT binding with a Hill number between 1 and 2 (data not shown). W84L or D313N mutation shifted the curve of CFT binding as a function of [Na+] to the left (Fig. 2a). The concentration for Na+ to produce half-maximal stimulation of [3H]CFT binding (EC50) was estimated by logistic analysis of the Na+-dependent binding (not including the specific binding in the absence of Na+) as a function of Na+ (up to where plateau binding was observed). At both W84L and D313N, Na+ displayed significantly lower EC50 values (Fig. 2), indicating an enhanced Na+ stimulation of CFT binding.

image

Figure 2. Biphasic effect of Na+ on [3H]CFT binding at wild-type (WT) and mutant hDATs. (a) 0–20 mm Na+ (b) 10–160 mm Na+. Cell membranes were incubated with 4 nm[3H]CFT for 15 min at 21°C in HEPES–Tris buffer containing various concentrations of NaCl. Results are shown as the percentage of the maximal binding activity. For wild-type, W84L, and D313N, the respective maximal binding to membranes was 2.81 ± 0.16, 3.38 ± 0.06, or 0.89 ± 0.13 pmol/mg. EC50 for Na+ is estimated by logistic analysis of the Na+-dependent binding [not including the specific binding in the absence of Na+ shown in (a)] as a function of Na+ (up to where plateau binding was observed). Values represent means ± SEM of more than eight experiments performed in triplicate. *p < 0.05 versus wild-type group (Dunnett's test).

Download figure to PowerPoint

In separate experiments, the effect of Na+ on [3H]CFT binding was tested at a higher range of concentrations (10–155 mm). As reported previously, higher concentrations of Na+, delivered either from NaCl (Fig. 2b) or Na+–ISE (data not shown), tended to reduce the [3H]CFT binding from its peak value. At high Na+ concentrations the relative binding level of CFT at W84L was not statistically significantly different from that at wild-type at any Na+ concentration higher than 20 mm. However, the CFT binding was significantly lower at D313N than at WT at 50 and 80 mm Na+ (Fig. 2b). At 155 mm Na+, [3H]CFT binding was reduced by about 33% at wild-type and D313N, and 20% at W84L, relative to the peak binding (Fig. 2b).

Na+ EC50 as a function of [CFT]

The stimulation by Na+ of CFT binding could be due to the possibility that CFT and Na+ prefer to bind to the same conformation(s) of the hDAT. If it is true, it follows from basic thermodynamic considerations that each ligand would increase the affinity of the hDAT for the other. Thus, at wild-type hDAT, EC50 for Na+ was assessed at varying concentrations of [3H]CFT and vice versa, Kd for CFT was assessed at varying concentrations of Na+. Indeed, raising [3H]CFT concentration to 16 nm significantly reduced the Na+'s EC50 (Fig. 3a), i.e. likely increased the apparent Na+ affinity, while raising Na+ concentrations reduced the CFT's Kd, i.e. increased the binding affinity for CFT (see next section).

image

Figure 3. Na+ EC50 for stimulating [3H]CFT binding as a function of [3H]CFT concentration (a) and the time curve for [3H]CFT binding in the absence of Na+ (b). Membranes of cells expressing wild-type hDATs were prepared in Na+-free HEPES–Tris buffer, and incubated with 4–16 nm[3H]CFT 21°C for 15 min in buffers containing various concentrations of NaCl (a) or for various time in the Na+-free buffer (b). Values represent means ± SEM of four experiments performed in triplicate. *p < 0.05 versus the value at 4 nm[3H]CFT (paired t-test).

Download figure to PowerPoint

Affinity of CFT and inhibitory potency of DA in the absence and presence of Na+

In the absence of Na+

[3H]CFT bound to membranes of HEK-293 cells expressing wild-type hDAT but not to that of HEK-293 cells expressing vector alone (Fig. 3b). This binding reached maximal level within 15 min at 21°C and remained at plateau for at least 45 min (Fig. 3b). Saturation analysis suggested a one-site model. Mutation of D313 significantly enhanced the binding affinity for CFT in the absence of Na+, while mutation of W84 did not (Table 1). The two mutants showed lower Bmax values than the wild-type hDAT, likely due to their modest expression level as indicated by immunostaining (Chen et al. 2001b). The apparent affinity for DA was assessed by measuring its apparent inhibitory constant (Ki) with respect to [3H]CFT binding. There were no significant differences in the inhibitory potency of DA between wild-type, W84L, and D313N (Table 1), suggesting that the mutations did not alter DA binding to the Na+-free hDAT.

Table 1.  [3H]CFT binding and inhibition by DA of [3H]CFT binding in Na+-free medium
hDATKd nmBmax pmol/mgDA Ki µm
  1. Data are expressed as means ± SE for 6–11 experiments performed in triplicate. ap < 0.05 versus wild-type (Dunnett's test).

Wild-type95.3 ± 9.511.3 ± 2.347.3 ± 1.07
W84L74.5 ± 7.76.96 ± 0.92a7.0 ± 1.60
D313N38.8 ± 5.4a4.14 ± 0.40a6.3 ± 0.89
In the presence of Na+

Na+ had no significant effect on Bmax values (Table 2). Raising Na+ concentration from 0 to 20 mm increased the binding affinity for CFT (Fig. 4a). We reasoned that if DA binds preferentially to the same hDAT conformation(s) as CFT, it should inhibit CFT binding in a similarly Na+-dependent manner. However, varying [Na+] had no effect on the Ki values for DA (Fig. 4a). No appreciable changes in DA Ki were observed even when the Na+–ISE concentration was raised to 155 mm.

Table 2.  Maximal binding capacities of [3H]CFT binding (pmol/mg) at various [Na+]
[Na+], mmWild-type hDATW84LD313N
  1. The Kd values of these experiments are shown in Fig. 4 and Table 5. Data are expressed as means ± SE for 4–11 experiments performed in triplicates. ap < 0.05 versus wild-type group at the corresponding [Na+]; no statistically significant differences between results at zero mm Na+ (see Table 1) and the Na+ levels tested within each group (Newman–Keuls test after log transformation of data).

  212.3 ± 1.26.41 ± 0.44a3.61 ± 0.40a
  412.2 ± 1.06.66 ± 0.41a3.57 ± 0.24a
 1011.0 ± 0.96.82 ± 0.38a3.37 ± 0.27a
 2013.4 ± 0.54.86 ± 0.74a2.80 ± 0.12a
15011.5 ± 0.46.98 ± 0.05a3.39 ± 0.07a
image

Figure 4. CFT affinity and DA inhibitory potency as a function of Na+ concentration. (a) Wild-type (WT); (b) W84L; (c) D313N. Cell membranes were incubated with 4 nm[3H]CFT for 15 min at 21°C in HEPES–Tris buffer containing various concentrations of Na+–ISE. Results are shown as the percentage of the binding affinity (1/Kd) or inhibitory potency (1/Ki) at zero mm Na+. For the Kd and Ki values at 0 and 20 mm Na+, see Tables 1 and 5. Values represent means ± SEM of six or more experiments performed in triplicate. *p < 0.05 versus the corresponding value for wild-type group (Dunnett's test).

Download figure to PowerPoint

If Na+ is not required for DA binding, it can be envisioned that a mutation primarily disturbs Na+ interaction with CFT without modifying the inhibition by DA of CFT binding. In support, the two mutations dramatically enhanced the stimulatory effect of Na+ on CFT affinity, mostly the plateau of the Na+ curve for W84L (Fig. 4b) and primarily a left shift in the curve for D313N (Fig. 4c), but had little effect on the Na+-insensitive feature of DA Ki (Fig. 4b,c). Both mutants showed CFT Kd values more than fivefold lower than wild-type hDAT at any Na+ concentrations tested. Thus, the mutations do not appear to simply promote Na+ binding. More likely, other mechanisms are involved, such as an underlying enhancement of the positive allosteric coupling between Na+ binding and CFT binding. In contrast, there were no significant differences in DA Ki values between wild-type and the mutants at any Na+ concentration tested (Fig. 4), suggesting that the mutations did not alter DA binding to the Na+-bound hDAT.

Inhibition by Li+ and K+ of [3H]CFT binding in the absence and presence of Na+

In the following experiments, we tested whether Na+ could impact CFT binding indirectly via interacting with other inhibitory cations such as Li+ and K+ at the transporter level. If it does, the mutations enhancing Na+ action could enhance such interactions, too. Thus, the interactions of Li+ and K+ with Na+ at the wild-type and mutated hDATs were assessed by testing the inhibitory effect of Li+ or K+ on [3H]CFT binding in the absence and presence of Na+.

In the absence of Na+

Under the Na+-free condition, there were no significant differences in the inhibitory potency for Li+(Table 3) and K+(Table 4) between wild-type, W84L, and D313N. Thus, mutations did not alter the interaction of the two cations with [3H]CFT sites at the Na+-free hDAT.

Table 3.  Potency of Li+ in inhibiting [3H]CFT binding at various [Na+]
Potency, mm[Na+], mmWild-typeW84LD313N
  1. Data are expressed as means ± SE for 3–4 experiments performed in triplicate. ap < 0.05 versus the wild-type group; bp < 0.05 versus the value at zero mM Na+ (Newman–Keuls test).

IC50 022.9 ± 2.736.9 ± 6.331.6 ± 3.0
 513.6 ± 0.353.3 ± 8.5a 87 ± 8a,b
2042.6 ± 4.2b149 ± 14a,b168 ± 18a,b
Ki 021.9 ± 2.635.0 ± 5.928.7 ± 2.7
 5 9.0 ± 0.3b14.9 ± 2.4a,b21.8 ± 2.0a
2032.9 ± 3.2b53.1 ± 4.9a72.3 ± 7.5a,b
Table 4.  Potency of K+ in inhibiting [3H]CFT binding at various [Na+]
Potency, mm[Na+], mmWild-typeW84LD313N
  1. Data are expressed as means ± SE for 3–4 experiments performed in triplicate. ap < 0.05 versus the wild-type group; bp < 0.05 versus the value at zero mM Na+ (Newman–Keuls test).

IC50 021.9 ± 2.018.9 ± 1.224.0 ± 2.4
 58.13 ± 0.9b19.2 ± 0.4a25.9 ± 2.2a
2030.6 ± 0.5b77.6 ± 4.9a,b84.3 ± 3.0a,b
Ki 021.0 ± 2.017.9 ± 1.221.8 ± 2.7
 55.37 ± 0.6b5.38 ± 0.1b6.46 ± 0.6b
2023.6 ± 0.427.7 ± 1.7b36.4 ± 1.3a,b
In the presence of Na+

Intriguingly, in the presence of 5 mm Na+ corresponding to the EC50 for stimulating [3H]CFT binding on the wild-type hDAT, the IC50 value for Li+ or K+ to inhibit [3H]CFT binding at wild-type hDAT was reduced (Tables 3 and 4). This reduction was not observed in the mutants (Tables 3 and 4). Li+, and also K+, by themselves, appeared to inhibit CFT binding in a competitive manner as judged from changes in Kd without changes in Bmax (data not shown). Thus, the IC50 values for the inhibitory cation would be elevated if the CFT binding affinity in the presence of Na+ was high enough to counteract the inhibitory effect of Li+ (or K+) at the CFT site. We therefore also computed the apparent Ki values for Li+ (and K+) by taking account of the binding affinity of CFT at each [Na+]. The Ki value for Li+ was significantly reduced by 5 mm Na+ in wild-type and W84L, but not in D313N (Table 3). The Ki value for K+ was reduced by 5 mm Na+ in all three hDATs (Table 4).

In contrast, further raising [Na+] to 20 mm increased both IC50 and Ki values for Li+ to the level above that in Na+-free medium and this increase was more profound in the two mutants (Table 3). At 20 mm Na+, the inhibitory potency of K+ was significantly weaker than that at 5 mm Na+ in both wild-type and mutant hDATs (p < 0.05, Table 4). However, compared with that in Na+-free buffer, the Ki values for K+ showed no change in wild-type and a slight increase in W84L and D313N (Table 4).

Taken together, the data from both wild-type and mutated hDATs are in agreement with an impact of Na+–Li+ interactions on CFT binding, in addition to the direct effect of the cations themselves. The different sensitivity to 20 mm Na+ in inhibiting CFT binding between Li+ and K+ and the relatively modest effect of the mutations on the Na+–K+ interaction indicate that the mechanism for Na+ to interact with K+ might be different from that with Li+.

Effect of Na+–Li+ interactions on the affinity of CFT and the inhibitory potency of DA

To address whether the Na+–Li+ interactions could impact DA binding differentially, we examined the effect of Na+–Li+ interactions on both the binding affinity of CFT and the potency of DA in inhibiting CFT binding. The CFT Kd and DA Ki were determined in the absence and presence of 60 mm Li+ at two fixed concentrations of Na+ (20 and 150 mm; Table 5). The two Na+ concentrations were chosen to allow a large change in Na+ concentration with limited change in the CFT affinity in the absence of Li+. Thus, if raising [Na+] reduces the Li+ inhibition, it should be mainly due to the antagonism by Na+ against Li+ action, rather than indirectly through enhancing CFT affinity. This reasoning also applies to DA binding, because DA Ki was Na+-independent (Fig. 4).

Table 5.  Effect of antagonism between Na+ and Li+ on the affinity for CFT and the potency for DA to inhibit CFT binding
  [Na+] 20 mm[Na+] 155 mm
Control[Li+] 60 mm% of controlControl[Li+] 60 mm% of control
  1. Data are expressed as means ± SE for 4–8 experiments performed in triplicate. ap < 0.05 versus wild-type group; bp < 0.05 versus control group at corresponding [Na+]; cp < 0.05 versus% of control values at 20 mm Na+ (Newman–Keuls test after log transformation of data).

CFT Kd, nmWild-type11.6 ± 2.7837.7 ± 4.95b324 ± 439.17 ± 1.5029.2 ± 4.09b327 ± 28
W84L2.22 ± 0.433.62 ± 0.70163 ± 32a1.48 ± 0.154.85 ± 1.44b310 ± 68
D313N2.45 ± 0.313.06 ± 0.14125 ± 6a1.69 ± 0.125.56 ± 0.92b338 ± 70c
DA Ki, µmWild-type6.27 ± 0.4922.8 ± 2.18b367 ± 506.42 ± 0.239.77 ± 0.80b153 ± 14c
W84L5.04 ± 0.4622.7 ± 2.99b466 ± 894.34 ± 0.375.20 ± 0.54119 ± 4c
 D313N4.25 ± 0.4016.4 ± 1.83b410 ± 725.11 ± 0.269.70 ± 0.50b190 ± 5c

Li+ (60 mm) displayed an inhibitory effect on the Kd of CFT binding (Table 5) but not its Bmax (data not shown). At 20 mm Na+, the percentage inhibition exerted by Li+ (60 mm) on CFT affinity was markedly less in the two mutants than in the wild-type hDAT (Table 5), which was consistent with the greater suppression of the Li+ effect by increasing Na+ from 5 mm to 20 mm in the two mutants (Table 3). Surprisingly, no further antagonism between Na+ and Li+ was observed at both wild-type and the mutants when [Na+] was raised from 20 mm to 155 mm (Table 5), though this antagonism was observed when [Na+] was raised from 5 mm to 20 mm, as judged from the increase in Li+'s Ki values (Table 3). Thus, a mutually exclusive interaction between Na+ and Li+ at the hDAT seems unlikely. Li+ also substantially increased DA Ki for inhibiting [3H]CFT binding, but this effect was considerably reduced at 155 mm Na+ (Table 5). The Na+ antagonism against Li+ inhibition of DA Ki was similar between wild-type and the mutants, arguing against an enhanced Na+ binding or a reduced Li+ binding at W84L and D313N.

Inhibition by H+ of CFT binding in the absence and presence of Na+

In addition to modify the effect of alkali cations (such as Li+ or K+) on the transporter function, Na+ may also play a role in deciding the pH sensitivity of the transporter. If this role is specific, a mutation-induced change in Na+ sensitivity could also be accompanied with a change in pH sensitivity. Thus, the effect of H+ on [3H]CFT binding was examined in the absence and presence of Na+.

In the absence of Na+

Reducing pH from 7.4 to 6.4 strongly inhibited [3H]CFT binding. The fractional decrease in CFT binding between wild-type and the mutants was similar (Fig. 5a), suggesting that the mutations did not markedly modify the interaction of H+ with [3H]CFT labeled sites at the Na+-free hDAT.

image

Figure 5. Inhibition by H+ of [3H]CFT binding in the absence (a) and presence (b) of Na+. Cell membranes were incubated with 4 nm[3H]CFT for 15 min at 21°C in Na+-free buffer or buffer containing 155 mm NaCl. In (a) the binding at pH 6.4 is expressed as percentage of that at pH 7.4. In (b) the binding is normalized to the maximal value at pH 9 and the pH 50 is estimated by logistic analysis of the binding as a function of pH. Values represent means ± SEM of 4–6 experiments performed in triplicate. *p < 0.05 versus wild-type group (Dunnett's test).

Download figure to PowerPoint

In the presence of Na+

With 155 mm Na+ in assay buffer, [3H] CFT binding peaked at pH 9 in both wild-type and the mutants. Thus, binding activities were normalized to the maximal value at pH 9 and the pH value producing 50% inhibition of the binding (pH50) were estimated by logistic analysis of the binding as a function of pH. The mutation shifted the curve of [3H]CFT binding as a function of pH to the right, with pH50 being significantly lower (Fig. 5b), indicating either a stronger Na+ antagonism against H+ inhibition or a higher affinity of CFT binding offsetting H+ inhibition.

Further analysis showed that Na+ did not simply oppose H+ effect. As shown in Fig. 6, when measuring [3H]CFT binding at pH 6.4 as percentage of the value at pH 7.4, Na+ showed biphasic effect on H+ inhibition of the binding, enhancing the inhibition at low concentrations and reducing it at higher concentrations (Fig. 6). Protection by Na+ from H+ inhibition can be calculated as the percentage increase in CFT binding over values without Na+. This protection was modest at wild-type hDAT, but significantly greater at the two mutants (Fig. 6). For wild-type hDAT, even at the saturating concentration of 155 mm, Na+ only partly counteracted the effect of H+ (15%). For W84L, Na+ protection started to occur at 40 mm (16%) and became greater at 155 mm Na+ (49%). For D313N, significant Na+ protection was observed at 10 mm and complete protection at 155 mm (Fig. 6).

image

Figure 6. Biphasic effect of Na+ on H+ inhibition of [3H]CFT binding at wild-type (WT) and mutant hDATs. Cell membranes were incubated with 4 nm[3H]CFT for 15 min at 21°C in HEPES–Tris buffer containing varying concentrations of NaCl at pH 6.4 and 7.4. Results are shown as percentage of the binding value at pH 7.4. Values represent means ± SEM of six experiments performed in triplicate. *p < 0.05 versus the respective group at zero mm Na+ (Dunnett's test).

Download figure to PowerPoint

When the Na+-dependent CFT binding data are expressed as percentage of the maximal binding at pH 7.4, H+ shifted the curve of [3H]CFT binding as a function of [Na+] to the right with reduced maximal binding (Fig. 7). This suggests allosteric interactions of H+ with Na+, affecting both the rising phase and plateau in the Na+ curve describing stimulation of [3H]CFT binding. At the two mutants, the Na+ EC50 values were more affected (+ 348 ± 17% for wild-type hDAT, + 880 ± 54% for W84L, and + 739 ± 26% for D313, p < 0.05 vs. wild-type group), but the plateau for Na+ was less affected (− 60.8 ± 0.55 for wild-type hDAT, − 33.8 ± 3.26 for W84L, and −12.2 ± 3.36% for D313N, p < 0.05 vs. wild-type group) by lowering pH (Fig. 7). Thus, the data support neither a weaker H+ binding nor a stronger Na+ binding, but indicate a tighter relationship between Na+ and CFT that hinders the effect of H+ on CFT binding.

image

Figure 7. Effect of lowering pH on Na+ stimulation of [3H]CFT binding. (a) Wild-type (WT); (b) W84L, (c) D313N. Cell membranes were incubated with 4 nm[3H]CFT for 15 min at 21°C in HEPES–Tris buffer containing various concentration of NaCl at pH 6.4 and 7.4. Specific [3H]CFT binding in the absence of Na+ was not included. Results are shown as percentage of the maximal level of Na+-dependent [3H]CFT binding at pH 7.4 (2.60 ± 0.12, 3.49 ± 0.10, and 0.77 ± 0.05 pmol/mg for wild-type, W84L, and D313N, respectively). Values represent means ± SEM of six experiments performed in triplicate.

Download figure to PowerPoint

Effect of Na+–H+ interactions on the affinity of CFT and the inhibitory potency of DA

In the following experiments, we tested whether Na+–H+ interactions, like Na+–Li+ interactions, could differentially influence CFT Kd and DA Ki. We determined the effect of one unit shift in pH on the CFT Kd and DA Ki at two fixed concentrations of Na+ (5 and 155 mm; Fig. 8). Again, the two Na+ concentrations allowed a broad comparison of Na+ concentration with limited change in the CFT Kd or DA Ki. Thus, if raising [Na+] reduces the H+ inhibition, it should be mainly due to antagonism between H+ and Na+, rather than indirectly through promoting CFT or DA binding.

For wild-type hDAT, lowering pH to 6.4 at 5 mm Na+ severely damaged CFT binding, as shown by the 35-fold increase in Kd (Fig. 8). Increasing [Na+] to 155 mm significantly attenuated the H+-induced inhibition at wild-type hDAT. For the mutants, at 5 mm Na+, the CFT affinity was also much lower at pH 6.4 (30-fold at W84L and 15-fold at D313N) than at pH 7.4. However, raising [Na+] to 155 mm produced a significantly greater protection at the mutants than at wild-type hDAT (Fig. 8).

For wild-type hDAT, there was no difference in the H+-induced increase in DA Ki between low [Na+] and high [Na+] (Fig. 8), suggesting that Na+ cannot prevent DA binding from H+ inhibition. This also ruled out the possibility that Na+ and H+ interact with hDAT in a mutually exclusive manner. However, increasing [Na+] to 155 mm almost completely prevented H+ from reducing DA Ki for the mutants (Fig. 8).

Discussion

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements

The current study was performed to address the possibility of CFT and DA binding to different hDAT conformations that are differentially impacted by Na+ and by other cations the binding of which can be affected by Na+. For this purpose, cationic interactions at the hDAT were explored in parallel. When interpreting cation interactions from their effects on ligand binding at the transporter, one faces multiple possibilities. First, Na+ and Li+ (or H+) can be thought to interact with hDAT mutually exclusively either via competing for the same site or inducing opposite conformational changes. In the present study, mutation at W84 or D313 specifically enhanced Na+ sensitivity (Fig. 2) but retained near normal Li+, K+, and H+ sensitivity in the absence of Na+ (Tables 3 and 4, Fig. 5a), suggesting that the sites at which Na+ exerts a stimulatory effect differ from those at which the inhibitory cations Li+, K+, or H+ act. Second, if Na+ and Li+ (or Na+ and H+) bind to different DAT conformations and the binding of each fully discourage the binding of the other (extreme negative allosterism resulting in mutually exclusivity), Na+ would prevent Li+ (or H+) from inhibiting not only CFT but also DA binding, even though the binding of DA itself was not sensitive to Na+. However, in wild-type hDAT, Na+ failed to completely overcome Li+ (or H+) inhibition of CFT binding and did not antagonize H+ inhibition of DA binding at all. Moreover, the inhibition by K+, Li+, and H+ of CFT binding was actually stronger at 5 mm Na+ than in Na+-free medium (Tables 3 and 4, Fig. 6). These findings do not support exclusive interactions between Na+ and Li+ (or H+). A third possibility that can be contemplated is that Na+ and Li+ (or H+) bind to hDAT simultaneously and do not interact with each other, but both interact with a third party on the hDAT. In this case, the apparent antagonism by Na+ of Li+ (or H+) inhibition as judged from [3H]CFT binding may be primarily due to the high affinity of CFT binding induced by Na+, which offsets the cation inhibition. However, Na+-induced reduction in Li+ inhibition remained even after correcting for CFT affinity (Table 3) or even when the Ki for DA was insensitive to [Na+] (Table 5). Additionally, Na+ antagonism against H+ inhibition was also observed when the control affinity for CFT was comparable between the two Na+ concentrations tested (Fig. 8). More likely, Na+ and Li+ (or H+) bind to the hDAT simultaneously, but the Na+ binding induces conformational changes that alter the interactions of Li+ (or H+) with the hDAT.

As previously reported (Reith and Coffey 1993; Corera et al. 2000), high concentrations of Na+ modestly reduced the CFT binding to a level below its peak stimulatory effect (Fig. 2b). Intriguingly, W84L mutation only shifted the rising phase while D313N mutation shifted both the rising and the declining phases of the Na+ curve to the left (Fig. 2). These observations raise the possibility that the dual effects of Na+ may represent different hDAT states triggered by Na+ binding at different sites of the transporter. Moreover, when comparing the inhibitory effect of Li+, K+, and H+ on CFT binding in the absence and presence of Na+, we found that Na+ influenced their effects also in a biphasic fashion, enhancing them at low concentrations and reducing them at moderate to high concentrations (Tables 3 and 4, Fig. 6). Such a complexity may reflect the contribution of more than one Na+-bound hDAT state to the mix of conformations existing at Na+ concentrations varying from 0 to 155 mm, causing different susceptibility to the inhibitory cations. All together, the results do not seem compatible with the idea that there is one hDAT conformation in the absence of Na+, another one in the presence of Na+, with [Na+] simply determining the proportion of transporters in each of those two states, and with Na+ counteracting the inhibitory action of Li+, K+, and H+ at the Na+-bound state. Rather, the results for membrane preparations under the current conditions suggest the existence of at least one more Na+ bound state where Na+ in fact enhances the inhibitory effect of other cations; this was not previously measured in other studies because the buffers commonly used do not take into account the 0–10 mm Na+ concentration range. Our experiments cannot answer the question of whether the states correspond to DAT free of Na+, DAT bound with one Na+ ion, and DAT bound with two Na+ ions (as during transport of DA). As observed for the allosteric effect of scorpion toxin on batrachotoxin binding to sodium channels (Catterall et al. 1985), if the transition between different hDAT conformations is very fast, one would actually measure a single time-averaged affinity for CFT depending on Na+ concentration. Therefore, the effect of the inhibitory cations would be expected to reflect their preference for a given hDAT state as well as the manner they interact with Na+ at that state.

In the present study, several lines of evidence from wild-type and mutated hDATs support the existence of binding conformations for DA that are distinguishable from those for CFT. Whereas Na+ markedly increased the transporter's affinity for CFT, it had no effect on the potency for DA to inhibit CFT binding. The observations that Na+ and CFT reciprocally enhanced each other's interaction with hDAT (Fig. 3a) suggest that Na+ and CFT bind to different sites and binding of each drive the protein towards a conformation favored by the other (coupling by positive allosterism). The lack of Na+ dependence of DA Ki observed here is consistent with the poor fits found for models assuming low DA binding to membrane preparations in the absence of Na+ (Li and Reith 2000), and with the observed lack of Na+ stimulation of DA binding with sulfate as the covarying anion (Li et al. 2002). The mutagenesis data further highlight the differences in Na+ dependence between CFT and DA binding. Thus, W84 or D313 mutation significantly enhanced the Na+ stimulation of CFT binding as well as the high-affinity CFT binding, but failed to modify the Na+-insensitive feature of DA Ki (Fig. 4). We conclude that CFT prefers a Na+-bound hDAT conformation(s), whereas DA can bind to Na+-free and Na+-bound hDAT conformations equally well, at least in membrane preparations. As a carrier transporting its substrate bidirectionally, the hDAT is expected to have multiple conformational transitions during the transport cycle. Thus, DA would have the ability to interact with multiple hDAT states where the DA binding sites face either high extracellular [Na+] or low intracellular [Na+]. In this context, membrane preparations may provide an opportunity to study conformational changes that cannot be readily detected in intact cells due to the restricted sidedness and cationic accessibility. However, more work is needed to characterize the role of Na+ in the binding of cocaine analogs and DA in functional plasma membranes of intact cells.

For CFT binding at wild-type hDAT, modest antagonism by Na+ of Li+ inhibition was observed when raising [Na+] from 5 mm to 20 mm (Table 3), but this effect was not further enhanced by raising [Na+] from 20 mm to 155 mm (Table 5). By contrast, Na+-induced reduction in Li+ inhibition of DA binding was observed when [Na+] was increased from 20 mm to 155 mm (Table 5). It can be considered that the proportion of various Na+-bound states of hDAT differs between moderate and saturating [Na+]. Further, on different and even on the same Na+-bound hDATs, the residues contributing to formation of CFT and DA binding sites are not identical. Thus, Li+ may have less of an effect on the formation of CFT binding sites on an intermediate hDAT state induced by modest [Na+] than on an hDAT state held by saturating [Na+]; the opposite could apply to the formation of DA binding sites. The findings that W84L or D313N mutation reduced Li+ inhibition of CFT binding at moderate but not high [Na+] as well as the observed lack of effect of Na+ antagonism against Li+ inhibition of DA binding (Table 5) lend further support to the above-mentioned ideas.

The discrepancy between CFT and DA binding was also evident when the effect of Na+–H+ interactions were examined in wild-type hDAT. At low [Na+], CFT Kd was much more sensitive to H+ than DA Ki; further, raising [Na+] attenuated H+ inhibition on CFT Kd but not DA Ki (Fig. 8a,d). Protonation of some critical residues may alter the charge-carrying status of the transporter, which not only directly impedes ligand binding, but also affects allosteric coupling between Na+ and CFT binding. Indeed, H+ lowered the plateau in the Na+ stimulation curve for CFT binding (Fig. 7), suggesting a less efficient Na+–CFT coupling. This could explain why CFT binding is more vulnerable to H+ than DA binding at low [Na+] and why increasing [Na+] cannot completely overcome the H+ inhibition of CFT Kd. In favor of a less efficient Na+–CFT coupling resulting from H+ action, the CFT binding affinity for W84L and D313N mutants that allowed higher Na+ plateau in stimulating CFT binding at low pH (Fig. 7b,c) was significantly tolerant to H+ inhibition (Fig. 8b,c).

In addition to cations inducing conformational changes in DAT, such as Na+ promoting a state preferentially binding cocaine-like compounds over DA, one also needs to consider the impact of ligand binding itself. For example, the thermodynamic analysis of Bonnet et al. (1990) indicated more drastic global conformational changes in the DAT upon binding of blockers than substrates, and Do-Régo et al. 1998) reported an early step in blocker binding that represented a conformational change. In addition, Ferrer and Javitch (1998) demonstrated that cocaine enhanced accessibility of C90 in hDAT to charged methanethiosulfonate (MTS) reagents, whereas benztropine was found to be inert in this respect as opposed to other blockers and DA itself (Reith et al. 2001). Another substrate, m-tyramine was reported to enhance accessibility to MTS of C342 in hDAT-expressing cells, requiring cocaine-sensitive translocation (Chen et al. 2000). Thus, cocaine analogs and DA may well induce different conformational changes in DAT, and it is very possible that these distinct conformational states are differentially sensitive to cations such as Li+ and H+. In this view, in the presence of saturating [Na+], rather than having one Na+-bound DAT state that binds CFT or DA with Li+ or H+ impacting binding of the two ligands differentially through non-identical intermediaries, one would have a state for the CFT–DAT complex that is different from the state of the DA–DAT complex.

The smaller EC50 value for Na+ stimulation of CFT binding at W84L and D313N could indicate an enhanced Na+-binding to Na+-preferring conformations. However, a smaller EC50 value could also result from kinetic changes in equilibrium binding assays as a consequence of enhanced coupling by positive allosterism between Na+ and CFT binding. Notably, the W84L mutation caused, in addition to a smaller EC50 value, a higher plateau in the Na+ stimulation curve for CFT affinity (Fig. 4), which indicates a stronger positive allosteric coupling effect. Further, for both mutants, the wild-type like Na+ antagonism against Li+ inhibition of DA binding (Table 5) as well as the greater H+-induced shift in Na+'s EC50 (Fig. 7) argue that the mutations neither enhanced the binding of Na+ nor reduced the binding of Li+ (or H+) in the presence of Na+. We consider it more likely that the mutations alter a Na+-bound form(s) of the hDAT. This alteration preferentially strengthens the positive allosteric coupling between Na+ binding and CFT binding, resulting in a shift in the binding selectivity of the mutants towards CFT over DA. Such an enhanced Na+–CFT coupling may also make CFT binding less sensitive to Li+ action (at modest [Na+]) and to the uncoupling effect of H+. In addition, there could be, even when [Na+] is saturating, multiple Na+-bound states with different affinities for CFT, giving an average plateau Kd; in this case, the mutations could cause increased frequency (and/or life-span) of a Na+-bound hDAT state(s) with higher affinity for CFT. Such changes would also explain the increased plateau in the Na+ curve for CFT affinity for W84L.

Although the mutations generally do not seem to affect DA binding in membrane preparations, they converted the H+ inhibition of DA Ki from being Na+-insensitive to Na+-sensitive (Fig. 8). This change could be accounted for by a Na+-bound form of the transporter less mobile in adopting H+-induced modification in conformational state or/and charge-carrying status that perturbs DA binding. We have also found that the DA uptake is damaged by the mutation of W84 and D313 (Chen et al. 2001b). How the altered cationic interactions occurring at recognition steps influence the transport cycle is currently under investigation. Because of the high conservation of W84 and D313 among the members of Na+/Cl-dependent neurotransmitter transporter family, the present findings may apply to other transporters. Indeed, mutation of a tryptophan residue equivalent to W84 of the hDAT in the GABA transporter GAT1 (W68L) enhanced the charge movement linked to Na+ binding (Mager et al. 1996), favoring a role of this residue common for the Na+/Cl-dependent transporter family.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements

Acknowledgements

We thank Dr Jonathan A. Javitch for the gifts of the pCIN4 vector and the hDAT cDNA; and the National Institute on Drug Abuse for supports (DA 08379 and 13261).

References

  • Xu C. and Reith M. E. A. (1996) Modeling of the pH dependence of the binding of WIN 35,428 to the dopamine transporter in rat striatal membranes: Is the bioactive form positively charged or neutral? J. Pharmacol. Exp. Ther. 278, 13401348.