Importance of Valine at Position 152 for the Substrate Transport and 2β-Carbomethoxy-3β-(4-fluorophenyl)tropane Binding of Dopamine Transporter

Materials and Methods

Site-directed Mutagenesis.

As described in the previous study (Lee et al., 1998), the eukaryotic expression vector pcDNA3.1(+) (InVitroGen, San Diego, CA)-based chimera hb3 has been constructed. It contains the chimeric DAT cDNA in which the region encompassing the third TM (from residues 133 through 186) of bovine DAT is substituted into human DAT. Seven of 54 amino acids in this region are different between human and bovine DATs. Each of these seven amino acid residues in chimera hb3 was transformed back to that of human sequence by site-directed mutagenesis. Site-directed mutagenesis was carried out using the QuickChange system (Stratagene, LaJolla, CA) against the 644-base pair (bp) HindIII-ScaI DNA fragment of DAT cDNA. Mutation-containing cDNA fragment was introduced into chimera hb3 cDNA by restriction enzyme digestion-ligation reaction, and the cDNA sequence was confirmed using an automatic sequencing system (Cy5 Autoread; Amersham Pharmacia Biotech, Uppsala, Sweden). The human DAT mutants, in which one of the amino acids at these seven positions was substituted by the corresponding bovine sequences, were also prepared using the same method. In this paper, the term Chi hb3 (I152V), for example, refers to the mutation of amino acid 152, isoleucine to valine in chimera hb3, whereas hDAT (V152I) designates the mutation on residue 152 of the human DAT from valine to isoleucine.

Functional Assay of DAT.

Wild-type and each of the mutated DAT cDNAs subcloned in the expression plasmid pcDNA3.1(+), were transfected into COS-7 cells using FuGENE6 reagent following the manufacturer's instructions (Boehringer Mannheim Co.). The cells were plated on 24-well plates with the number of 3 × 10⁴ cells per well, and grown overnight in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum. DNA (0.3 μg/well) was mixed with FuGENE6 reagent (0.5 μl/well) and incubated for 15 min at room temperature. The DNA·FuGENE6 complex was added on the cells. After incubating for 2 days, [³H]DA uptake, [³H]MPP⁺ uptake, and the tritium-labeled cocaine analog 2β-carbomethoxy-3β-(4-fluorophenyl)tropane ([³H]CFT, also designated as [³H]WIN35,428) binding assays were carried out in a modified Krebs-Ringer HEPES (KRH) buffer containing 25 mM HEPES (pH 7.4), 125 mM NaCl, 4.8 mM KCl, 1.2 mM KH₂PO₄, 1.3 mM CaCl₂, 1.2 mM MgSO₄, 5.6 mM glucose, 1 mM sodium ascorbate (for [³H]DA uptake, 10 μM pargyline was added). Cells were washed with the modified KRH buffer then incubated with 50 nM [³H]DA (51 Ci/mmol, Amersham Pharmacia Biotech) or 10 nM [³H]MPP⁺ (83 Ci/mmol, DuPont-New England Nuclear Co., Boston, MA) for 10 min at 37°C. The uptake reaction was terminated by adding ice-cold KRH buffer, and the cells were washed with the same buffer three times. The cells were solubilized in 0.5 M NaOH, and their radioactivity was analyzed using a liquid scintillation counter. The binding of [³H]CFT to DATs on intact cell membranes was determined by incubating cells with the KRH buffer containing 10 nM [³H]CFT (87 Ci/mmol, DuPont-New England Nuclear) for 2 h at 4°C. Specific uptake or binding were defined as the difference between the uptake or binding of the cells transfected with the cDNA for DATs and those of the cells transfected with the pcDNA3.1 vector alone.

For saturation analysis of DA uptake, the [³H]DA concentration was maintained as 100 nM and the unlabeled DA concentration was adjusted from 0.5 to 8 μM. In the case of MPP⁺ uptake and CFT binding, they were performed under the presence of 10 nM [³H]MPP⁺ and 10 nM [³H]CFT with varying concentrations of unlabeled MPP⁺ (0 to 10 μM) or CFT (0 to 0.1 μM), respectively. The estimated K_m andV_max values were obtained from the curve-fitting program of SigmaPlot (SPSS, Chicago, IL).

Confocal Microscopic Analysis.

pEGFP-C1 vector (CLONTECH, Palo Alto, CA), designated to express the cloned protein fused to enhanced green fluorescent protein (EGFP) (Prasher et al., 1992) at its N-terminal end, was used to identify the membrane localization of wild-type or mutated DAT proteins. pEGFP-DAT clones were constructed by insertion of wild-type or mutant DAT cDNAs into the multiple cloning site of pEGFP-C1 by HindIII-XbaI restriction and ligation reactions. They were transiently transfected into COS-7 cells as described above, and their localization was examined using a confocal microscope (Bio-Rad, Hercules, CA) with a microscope (Axiphot; Zeiss, Zena, Germany; magnifications: objective 20×, eyepiece 10×). To make a reference for the membrane localization of transporter protein, COS cells transfected with pEGFP-hDAT were treated with 200 nM phorbol 12-myristrate 13-acetate for 10 min, and the internalization of transporter protein was induced. Before confocal microscopic analyses, functional activities of EGFP-fused DAT, including DA uptake and CFT binding, were tested to confirm if they were functionally active.

Results

Recovery of DAT Functions in Chimera hb3 by Single Amino Acid Substitution.

Expression of wild-type human DAT subcloned into the expression plasmid pcDNA3.1(+) conferred avid uptake of DA and MPP⁺ and CFT binding activity in the COS cells transfected with the plasmid, as described earlier (Lee et al., 1996,1998). The substitution of the region encompassing the third TM (from residues 133 through 186) of bovine DAT for human DAT, as shown in chimera hb3 (Fig. 1), remarkably reduced the DA uptake, MPP⁺ uptake, and CFT binding to 5.4 ± 1.5, 2.4 ± 2.4, and 20.3 ± 5.5% of those of human DAT, respectively (Fig. 2).

To determine what amino acids are responsible for such marked reductions in DAT functions, several different amino acid residues of bovine sequence in the chimera hb3 were replaced back to the original human DAT sequence and followed by characterization of the DAT actions. Putative membrane topology of DAT as determined by hydrophobicity analyses (Uhl and Hartig, 1992; Amara and Kuhar, 1993; Giros and Caron, 1993) indicates that the region between the 133rd and 186th residues spans the first intracellular loop (amino acid positions from 133 through 142), the third TM (143 through 166), and the second extracellular loop (167 through 186) as shown in Fig. 1. When bovine sequence alanine at position 139 on the first intracellular loop of chimera hb3 was transformed back to human sequence lysine [Chi hb3 (A139K)], no change was observed on the transport and binding activities. The substrate transport and CFT binding were still as low as those observed in the chimera hb3 (Fig. 2). Substitutions of amino acids on the second extracellular loop (residues 168, 178, and 182) also did not affect DA and MPP⁺ uptake activities of the chimera hb3, although these substitutions resulted in slight increases in [³H]CFT binding activities. [³H]CFT binding was increased significantly by 15% upon substitution of asparagine for histidine at position 182 of chimera hb3.

Human and bovine DAT differ in their deduced primary structure within the third TM at three positions: tyrosine in bovine DAT instead of phenylalanine in human DAT at position 143, alanine instead of valine at position 145, and isoleucine instead of valine at position 152. When each of the residues at positions 143 and 145 of chimera hb3 was replaced back to the corresponding human DAT sequences [Chi hb3 (Y143F) and Chi hb3 (A145V)], the DAT activities still remained comparable with those of chimera hb3.

On the contrary, substitution of valine for isoleucine at residue 152 dramatically restored DA uptake, MPP⁺ uptake, and CFT binding to 84.3 ± 11.1, 76.9 ± 10.5, and 97.5 ± 30.8% of human DAT, suggesting that the amino acid at position 152 is responsible for the marked decreases of DAT functions in the chimera hb3. There is no significant difference inV_max and K_mvalues between hDAT and chimera hb3 (I152V), except that theK_m value for DA uptake was increased (Fig.6 and Table 1) in chimera hb3 (I152V). Confocal microscopic analyses on COS cells transfected with EGFP-hDAT or EGFP-Chi hb3 (I152V) showed no difference in their membrane localization (Fig. 3).

Loss of Functions in Human DAT by Single Amino Acid Substitution for Bovine Residues.

To confirm critical importance of the 152nd residue valine in conferring DAT functions, another set of mutant DATs, in which each of the seven amino acids in the region encompassing the third TM of human DAT was converted to the corresponding bovine sequence, was prepared and assessed for DA uptake, MPP⁺ uptake, and CFT binding (Fig.4). The mutant hDAT (V152I), in which valine at position 152 was replaced by isoleucine, showed remarkably reduced DA uptake, MPP⁺ uptake, and CFT binding activities to 15.9 ± 3.8, 2.6 ± 0.5, and 42.7 ± 11.1% of the human DAT, respectively (Fig. 4). However, it still showed a similar membrane localization pattern to wild-type hDAT when examined under a confocal microscope (Fig. 3).

Interestingly, the substitution of alanine for valine at position 145 revealed that MPP⁺ uptake was reduced to 4.2 ± 1.9% of human DAT value, whereas the DA uptake and the CFT binding did not change significantly from those of the wild-type human DAT. Slight reductions of the MPP⁺ uptake and CFT binding activities were detected by the substitution of residue 168. The conversion of the 182nd residue showed that the CFT binding of human DAT was significantly reduced by 20% without any change of DA and MPP⁺ uptake activities. Except for these activities, no significant change of substrate transport and CFT binding was observed by any conversions of amino acids at positions 139, 143, and 178 of the human DAT.

Substitutions of Various Amino Acids for Isoleucine at Position 152 of Chimera hb3.

To test whether DAT function would depend on the side chain size of amino acid 152 or not, isoleucine at position 152 of chimera hb3 was transformed into various monoamino, monocarboxylic amino acids. Transformation of isoleucine to alanine significantly increased all the functional activities of chimera hb3 by 16 to 55% (Fig. 5). But the effects were not as big as in the case of the conversion to valine, which showed greatly decreased V_max and increasedK_m values (Fig.6 and Table 1) compared with those for hDAT and chimera hb3 (I152V). On the confocal microscopic analysis (Fig. 3), however, chimera hb3 (I152A) showed a similar membrane localization pattern to hDAT and chimera hb3 (I152V).

By contrast, neither of the conversions to leucine nor glycine affected the functional activities of the chimera hb3. Isoleucine was also replaced by aspartic acid or histidine, of which sizes of side groups are similar to valine, to test whether the size occupied by the side group of the 152nd residue would be a determinant to the DAT functions. Neither of these transformations showed any effect on the functions of chimera hb3, suggesting that the size of amino acid at the position 152 may not be the most critical determinant for DAT functions.

Discussion

Although studies (Kitayama et al., 1992b, 1993; Buck and Amara, 1994, 1995; Giros et al., 1994; Wang et al., 1995; Pifl et al., 1996) have been undertaken to identify the underlying structural determinants for DAT function, we still lack complete knowledge of the specific residues contributing for DAT function. Like other members of this gene family (Jayanthi et al., 1998), the DAT amino acid sequence predicts 12 TMs, a large extracellular loop between the third and fourth TMs, and multiple consensus sequences in the intracellular domains for possible phosphorylation by protein kinases. Although inconsistencies have been reported for the 12-TM model (Bennet and Kanner, 1997; Clark, 1997), the predicted 12-TM topology for DAT has been supported by biochemical and mutagenesis studies (Buck and Amara, 1994; Wang et al., 1995). In this study, based on the finding that the DAT region encompassing the third TM is important for function, we have tried to determine the specific amino acid(s) contributing to the substrate transport and CFT binding in this region. The activities for DA uptake, MPP⁺ uptake, and CFT binding remarkably reduced only by substituting the third TM-containing region of the bovine DAT (from residues 133 through 186) into the human DAT (chimera hb3). Although there is a possibility that the structural perturbation of the chimera hb3 have affected an indirect impact on regions elsewhere in the transporter protein, it is hypothesized that one or more amino acids in this region would be primarily responsible for this loss of function. In the case of serotonin transporter, it was also suggested that the third TM contains residues associated with substrate and cocaine analog binding (Chen et al., 1997). Comparing the amino acid sequences, only seven of 54 amino acids in this region are different between human and bovine DAT. We have interchanged each of these seven different amino acids using site-directed mutagenesis and tried to determine the amino acid(s) contributing to the functional importance of the region encompassing the third TM.

It has been suggested that the protonated amine of monoamines is likely to be associated with a negatively charged residue of the transporter, whereas planar aromatic moieties of these compounds are believed to be associated with an analogous surface of the transporter by hydrophobic and/or van der Waals bonding (Maxwell et al., 1976; DePaulis et al., 1978). Based on these hypotheses, it could be assumed that phenylalanine at residue 168 of human DAT would interact efficiently with the aromatic rings of DA, MPP⁺, and CFT, whereas leucine, residue 168 of chimera hb3, would not. The amino acids at positions 143 and 178 of human and bovine DAT are relatively different in their hydrophobicity: the 143rd amino acid is tyrosine in bovine DAT instead of phenylalanine in human DAT, and residue 178 is threonine instead of isoleucine. Considering that the hydrophobicity of amino acids is one of the most important factors determining the functional structures of protein, these amino acids could be responsible for the loss of function after substituting the third TM-containing region. However, the present results do not follow these theoretical assumptions. The interchanges between human and bovine amino acid sequences at positions 143, 168, and 178 did not influence the functional activities of the DAT, although the CFT binding was slightly affected by the substitution at residue 168. On the contrary, the substitution of valine for isoleucine at position 152 of chimera hb3 dramatically increased the substrate transport and CFT binding almost to the level of the human DAT with similarV_max and K_mvalues. Moreover, the functional activities of the human DAT decreased nearly to the values of chimera hb3 by substituting isoleucine for valine. These findings clarify that the 152nd residue, valine, is crucial for the substrate transport and CFT binding in human DAT.

Comparing the residues at position 162, isoleucine and valine, of chimera hb3 and the human DAT, both are monoamino, monocarboxylic amino acids whose side groups solely consist of a hydrocarbon chain. Their physicochemical natures, including their hydrophobicity, are also almost identical. Therefore, the current substitution is not likely to induce drastic alterations in the secondary or tertiary structure of DAT, but the possibility of this inducement cannot be excluded. Instead, it might be implicated that a specialized chemical group is involved in recognizing or translocating the substrates. Because the only reliable difference in these amino acids is the length of hydrocarbon chains or the size occupied by their side groups, we have substituted various amino acids with different sizes for valine or isoleucine. (The proposed van der Waals volumes of valine and isoleucine are 105 and 124 A³, respectively.) In this study, the conversion of isoleucine to glycine or leucine did not increase the functional activities of chimera hb3. We also converted isoleucine of chimera hb3 to aspartic acid or histidine whose sizes are similar to valine. The proposed van der Waals volume of glycine, leucine, histidine, and aspartate are 48, 124, 118, and 91 A³, respectively. Neither aspartic acid nor histidine at position 152 of the chimera hb3 recovered the function of DAT. On the contrary, when isoleucine was converted to alanine (van der Waals volume, 67 A³) at position 152, the functional activities of chimera hb3 were recovered to a significant level, although the values were not as high as those recovered by converting to valine. By saturation analyses of Chi hb3 (I152A) for DA uptake, MPP⁺ uptake, and CFT binding, theV_max values were significantly decreased and the K_m values were significantly increased compared with those of wild-type or Chi hb3 (I152V). These results suggest that the size of the residue at position 152 could not be the only factor specifying the substrate transport or CFT binding of DAT, even though it would be the most important one.

To clarify the role of valine at position 152, we tested whether the loss of function in chimera hb3 resulted from disturbed trafficking of DAT to the cell membrane. Under confocal microscope examination, no significant difference in the expression pattern was observed between the functionally active DATs [WT hDAT, Chi hb3 (I152V)], Chi hb3, Chi hb3 (I152A), and hDAT (V152I). This result might support the possibility that certain amino acids in the region encompassing the third TM participate in the direct interaction with substrates or at least in forming the substrate binding structure.

Depending on the studies of the G protein-coupled receptors, it is assumed that monoamine ligands bind to the TM core of receptors, which consists of three TMs (Gether and Kobilka, 1998; Ji et al., 1998). In the case of transporter, including DAT and SERT, the first, third, and seventh TMs have been known to be important for ligand interaction. Mutations of a conserved aspartate in the first TM (position 98 in rat SERT or position 79 in human DAT) have led to decreased substrate transport activity and even the selective loss of antagonist potencies, including cocaine analog (Kitayama et al., 1992; Barker et al., 1999). The amino acid serine at positions 356 and 359 in the seventh TM of DAT (Kitayama et al., 1992) and the isoleucine at positions 172 and 179 in the third TM of SERT (Chen et al., 1997) are postulated to interact with their monoamine ligands. Like the isoleucine in the third TM of SERT, it can be suggested that the valine at position 152 in the third TM of DAT might be involved in the direct interaction of DAT with its ligands. The amino acid valine is well conserved at the same position in the third TM of catecholamine transporters except bovine DAT (Fig.1), although it is isoleucine in serotonin transporter.

In addition to the functional importance of the 152nd residue, the present study also suggests that each of the DA uptake, MPP⁺ uptake, and CFT binding activities would be mediated by DAT in different manners, despite the possible existence of a common mechanism for all. Asparagine at position 182 and phenylalanine at position 168 might be involved in the CFT binding without affecting DA and MPP⁺ uptake, whereas valine at position 145 is possibly related to MPP⁺ uptake.

Determination of the mode of interaction of DA, MPP⁺, and cocaine on DAT is prerequisite for the fundamental solutions of DAT-relating disorders such as cocaine addiction, Parkinson's disease, and schizophrenia. As more information becomes available on the higher-order structure, the discrete molecular basis for the actions of DA, MPP⁺, and cocaine on DAT will become more apparent.