Mutational effects of human dopamine transporter at tyrosine88, lysine92, and histidine547 on basal and HIV-1 Tat-inhibited dopamine transport

Dysregulation of dopaminergic system induced by HIV-1 Tat protein-mediated direct inhibition of the dopamine transporter (DAT) has been implicated as a mediating factor of HIV-1 associated neurocognitive disorders. We have reported that single point mutations on human DAT (hDAT) at tyrosine88 (Y88F), lysine92 (K92M), and histidine547 (H547A) differentially regulate basal dopamine uptake but diminish Tat-induced inhibition of dopamine uptake by changing dopamine transport process. This study evaluated the effects of double (Y88F/H547A) and triple (Y88F/K92M/H547A) mutations on basal dopamine uptake, Tat-induced inhibition of DAT function, and dynamic transport process. Compared to wild-type hDAT, the Vmax values of [3H]Dopamine uptake were increased by 96% in Y88F/H547A but decreased by 97% in Y88F/K92M/H547A. [3H]WIN35,428 binding sites were not altered in Y88F/H547A but decreased in Y88F/K92M/H547A. Y88F/H547A mutant attenuated Tat-induced inhibition of dopamine uptake observed in wild-type hDAT. Y88F/H547A displayed an attenuation of zinc-augmented [3H]WIN35,428 binding, increased basal dopamine efflux, and reduced amphetamine-induced dopamine efflux, indicating this mutant alters transporter conformational transitions. These findings further demonstrate that both tyrosine88 and histidine547 on hDAT play a key role in stabilizing basal dopamine transport and Tat-DAT integration. This study provides mechanistic insights into developing small molecules to block multiple sites in DAT for Tat binding.

Using computational modeling and experimental approach, we have identified several key residues on human DAT (hDAT), which are crucial for Tat-hDAT interaction and dynamic transport process 9 . Furthermore, we have demonstrated that in vitro exposure to Tat reduced reuptake of DA via hDAT in cells [16][17][18] and rat striatal synaptosomes 19 . The inhibitory effect of Tat on DAT function results from Tat directly interacting with DAT 16,20,21 . Single point mutations of hDAT at tyrosine88 (to phenylalanine, Y88F), lysine 92 (to methionine, K92M), his-tidine547 (to alanine, H547A) differentially alter basal DA uptake but attenuate the Tat inhibitory effects on DA transport 17,18 . For example, DA uptake is decreased in K92M and increased in H547A, respectively; however Y88F mutant preserves basal DA uptake 17,18 . Notably, the mutational effects on normal DA uptake and Tat inhibitory effect on DAT function are associated with alterations of transporter conformational transitions 9 .
We have demonstrated that Tat protein inhibits DAT function in an allosteric modulatory mechanism 19,22 . Recent studies have demonstrated that novel SRI-compounds exhibit a partial antagonistic role in DAT function as allosteric modulators [23][24][25] . We have reported that SRI-30827, one of the novel allosteric modulators, blocks Tat interaction with DAT 26 . Thus, identifying the specific binding sites on hDAT for Tat and its role in DA transport process could be beneficial to attenuation of the inhibitory effect of Tat on DAT-mediated dopaminergic transmission. On the other hand, through an allosteric modulatory mechanism, the inhibitory effect of Tat on DAT function can also be diminished by targeting the specific DAT residues that are distinct from Tat binding sites. However, based on our computational structural models for Tat binding with hDAT, the interaction of Tat with hDAT involves multiple residues of DAT 9,21 and our previous results obtained from single point mutations of DAT only present the role of a particular residue in Tat-DAT interaction. Therefore, this study investigated the mutational effects of Y88F/H547A and Y88F/K92M/H547A on basal DAT function, Tat inhibitory effect on DA uptake, and dynamic DA transport process.

Results
Computational modeling: Impact of Tyr88/His547 and Tyr88/Lys92/His547 on functional relevance of human DAT. Based on the constructed hDAT-Tat binding model in our previous work 20 , hDAT residues Y88, K92, and H547 could form hydrogen bonds with residues K19, P18, and R49 of HIV-1 Tat, respectively. According to our previous computational and experimental results, single mutation of either hDAT Y88 (Y88F), K92 (K92M) or H547 (H547A) could significantly attenuate the binding between DAT and Tat 17,18 . As shown in Fig. 1A,B, hDAT Y88, K92 and H547 residues independently interact with Tat, suggesting an effect for double or triple mutation on disturbing DAT-Tat binding. For example, two hydrogen bonds in the DAT-Tat interface were eliminated by double mutation of Y88F/H547A on hDAT (Fig. 1C), and three hydrogen bonds in the DAT-Tat interface were eliminated by triple mutation of Y88F/K92M/H547A on hDAT (Fig. 1D). Therefore, one may expect that Y88F/H547A and Y88F/K92M/H547A mutants are more effective in inhibiting DAT-Tat binding than single point mutations of Y88F, K92M and H547A.
In the hDAT, the [ 3 H]WIN35,428 binding sites share pharmacological identity with the DA uptake site and is also part of the cocaine binding domain 27,28 . Kinetic analysis of [ 3 H]WIN35,428 binding was conducted to determine the effects of multiple mutations of hDAT on the DA binding site. As shown in Table 2

Mutation of Y88/H547 alters the inhibition potency of DA uptake and DAT binding by substrates and inhibitors.
To characterize the influence of Y88F/H547A on DA transport and the specific binding site in hDAT for DA, cocaine, GBR12909, and amphetamine (AMPH), we determined the ability of these DAT inhibitors or substrate to inhibit [ 3 H] DA uptake in WT hDAT and Y88F/H547A (Table 1). In comparison to WT hDAT, Y88F/H547A reduced the apparent affinity (IC 50 30 . We have demonstrated that the addition of 1 µM PMA produced a 40% and 60% decrease in the V max of DA uptake in WT hDAT and H547A mutant, respectively, compared to the respective control, suggesting that H547A displays increased sensitivity to PMA-induced DAT phosphorylation 18 . In the current study, we also found that Y88F/H547A displayed increased the V max relative to WT hDAT. We then determined whether Y88F/ H547A-upregulated DA uptake is also mediated by alteration of sensitivity to PMA-induced DAT phosphorylation (Fig. 3) (5) = 2.54, p < 0.05], respectively. These results suggest that the double mutation of Tyr88 and His547 attenuates the single mutant H547A-increased sensitivity of PMA-induced decrease in DA uptake.    17,18 . Since the triple mutant (Y88F/K92M/H547A) displayed an extremely low V max of [ 3 H]DA uptake, only the double mutant (Y88F/H547A) was tested for its effect on Tat-induced inhibition of DA uptake. Based on the computational modeling predictions (Fig. 1), the double mutation of Tyr88 and His547 would eliminate two pairs of hydrogen bonds between D-Y88-T-K19 and D-H547-T-R49 and impair Tat binding on hDAT, thereby attenuating Tat-induced inhibition of DA uptake. Due to the difference in the specific [ 3 H]DA uptake between WT hDAT and Y88F/H547A mutants (Fig. 2), the inhibitory effects of Tat on DAT function in WT and Y88F/H547A were presented as % of Tat-mediated [ 3 H]DA uptake relative to their respective   www.nature.com/scientificreports www.nature.com/scientificreports/ controls (in the absence of Tat). As shown in Fig. 4, the application of 140 nM recombinant Tat 1-86 (rTat 1-86 ) significantly reduced the specific [ 3 H] DA uptake in WT hDAT by 31% relative to its control [t (4) = 11.54, p < 0.001]. However, the Tat-induced reduction of [ 3 H] DA uptake in WT hDAT was attenuated in Y88F/H547A compared to its control [t (4) = 0.42, p > 0.05], suggesting that the double mutation of Tyr88 and His547 attenuates Tat's inhibitory effect on DA uptake.
Effects of double mutation of Y88/H547 on zinc regulation of DAT conformational transitions and basal DA efflux. Tat protein has been shown to regulate DA transport allosterically, which may lead to alteration of transporter conformational transitions 19,22 . We have previously demonstrated that His547 residue www.nature.com/scientificreports www.nature.com/scientificreports/ plays a critical role in allosteric modulation of Tat-DAT interaction 18 . To determine whether the His547 residue still acts as a potential site for the regulation of Tat-DAT in Y88F/H547A, we examined the effects of Y88F/ H547A on zinc modulation of [ 3 H]DA uptake and [ 3 H]WIN35,428 binding. Addition of Zn 2+ is able to partially reverse an inward-facing state to an outward-facing state 31,32 . On the basis of this principle, the addition of Zn 2+ to WT hDAT would inhibit DA uptake, whereas in a functional mutation of DAT, Zn 2+ might diminish the preference for the inward-facing conformation and thus enhance DA uptake. As shown in Fig Effects of Y88/H547 mutation on amphetamine-regulated DA efflux and uptake. AMPH acts as a substrate for DAT which subsequently reverses the transporter leading to DA efflux 33,34 . AMPH-stimulated [ 3 H] DA efflux in PC12 cells transfected with WT hDAT and its mutants was determined under a stable baseline of DA efflux (Fig. 6A), showing no difference on basal [ 3 H]DA efflux in control groups (F (4, 27) = 1.92, p = 0.14, one-way ANOVA). As shown in Fig. 6B  www.nature.com/scientificreports www.nature.com/scientificreports/   www.nature.com/scientificreports www.nature.com/scientificreports/ and Y88F/H547A in presence of AMPH (1 µM, final concentration). As shown in Fig. 6C, two-way ANOVA analysis on the V max of [ 3 H]DA uptake reveled significant main effects of mutation (F (1, 16) = 9.50, p < 0.01) and AMPH (F (1, 16) = 11.99, p < 0.01). No significant interaction of mutation × AMPH treatment was found (F (1, 16) = 1.61, p = 0.22). Post hoc analyses show that the V max value in Y88F/H547A (18.9 ± 2.81 pmol/min/10 5 cells) was higher than that in WT hDAT (11.5 ± 0.94 pmol/min/10 5 cells, t (8) = 2.49, p < 0.05, n = 5/group) in the absence of AMPH. AMPH decreased the V max in both WT hDAT (32%, 7.80 ± 0.97 pmol/min/10 5 cells, t (8) = 2.76, p < 0.01) and Y88F/H547A (42%, 10.9 ± 1.33 pmol/min/10 5 cells, t (8) = 2.58, p < 0.05) compared with their respective controls. There was a significant difference in the magnitude of the AMPH-induced decrease in V max values between WT hDAT and Y88F/H547A (t (8)

Discussion
Our recent studies demonstrated that single hDAT mutants, Y88F, K92M, and H547A differentially regulate DAT reuptake but attenuate Tat's inhibitory effect on DAT function. This study sought to determine whether multiple mutations on these hDAT residues (Y88F/H547A) and (Y88F/K92M/H547A) influence the direct interactions between Tat and DAT. There were three major findings. First, compared with WT hDAT, Y88F/H547A enhances V max of DA uptake, whereas Y88F/K92M/H547A dramatically reduces the V max . Second, consistent with the computational modeling prediction, Y88F/H547A attenuates Tat-induced inhibition of DA uptake observed in WT hDAT. The third finding is that Y88F/H547A displays an attenuation of zinc ion modulation of [ 3 H]WIN35,428 binding, increased basal DAT-mediated DA efflux, and reduced AMPH-stimulated DA efflux. Collectively, these findings suggest that targeting DAT residues Tyr88 and His547 may have potential therapeutic benefits on Tat-induced dysregulation of dopaminergic system seen in patients with HAND.
According to our computationally modeled structures, Y548-Y470-Y551 is a stable motif (denoted as the YYY motif for convenience) essential for the conformational conversion of hDAT required for the DA transport process 35 . This YYY motif has hydrophobic contact with a typical U-turn loop associated with extracellular loop 6 (EL6) and transmembrane 10 (TM10a). Any structural changes affecting the stability of the YYY motif could destabilize the R85-D476 salt bridge of hDAT, and the salt bridge is key for the conformational conversion [36][37][38][39] . H547 is expected to significantly affect the hDAT stability because H547 is adjacent to Y548 of the YYY motif 35 , which is consistent with the observed increase in DA uptake by the H547A mutation 18 . On the other hand, the side chain of K92 residue (TM1b) forms a favorable salt bridge with the side chain of D313 (TM6a) during the molecular dynamics simulation on the hDAT-DA binding complex, and the bridge is also essential for the conformational conversion of hDAT required for the DA uptake process 40 . For this reason, disrupting the K92-D313 salt bridge will likely impair or slow down the DA uptake process, which has been validated pharmacologically by mutations on K92 (K92M) 17 and D313 (D313N) 41 . In addition, the aromatic ring on the side chain of Y88 is sandwiched by TM1b and extracellular loop 4 (EL4) in a hydrophobic region of hDAT. Therefore, mutation of Y88 to phenylalanine retains the normal reuptake function of hDAT and demonstrated no difference compared to WT hDAT 17 . However, Y88F increased the K m of DA binding by about 50%, suggesting that Y88F may slightly alter the local structure around Y88 17 . In the current study, the double mutant Y88F/H547A retains the increased V max observed in H547A, suggesting a critical role of the mutated H547 residue in regulating transporter activity. In contrast, the combination of the mutated residues Y88, K92, and H547 does not follow a simple math addition, because compared to WT hDAT, a 96% reduction of V max was observed in Y88F/K92M/H547A, whereas the V max was decreased by 71% in K92M 17 . These results suggest that the inclusion of the K92 mutation in the K92-D313 salt bridge is a critical factor in the differential DAT kinetics in the triple mutation of hDAT. As Y88 is spatially adjacent to R85 and K92 in helix TM1b, the double mutations on both Y88F and K92M are expected to induce more structural disturbance to the R85-D476 and K92-D313 salt brides, thus further decreasing V max of the transporter. Therefore, triple mutant Y88F/K92M/H547A has a lower V max compared to, single mutant K92M. Although including K92M mutation significantly decreases the V max for DA uptake in the triple mutant Y88F/K92M/ H547A compared to Y88F/H547A, no significant change in the B max of [ 3 H]WIN 35,428 binding is found in the double or triple mutants, suggesting that the multiple mutations may alter DAT re-uptake function rather than the substrate binding sites on DAT. This conclusion is also supported by the increased K m in Y88F/H547A and no change in K m in Y88F/K92M/H547A. However, single point mutations of Y88, K92, and H547 differentially affect the V max with no changes in the K m 17,18 . One possible explanation is that allosteric modulation of the dynamic DA transport process may be responsible for the enhanced V max and its associated DA uptake potency.
We further determined the effect of double mutant Y88F/H547A on the basal DA uptake via hDAT and binding potency for DA, GBR12909, and cocaine. Y88F/H547A had decreased the potency for [ 3 H]DA to inhibit DA uptake relative to WT hDAT, which is similar to that in Y88F and H547A 16,18 . However, Y88F/H547A increased the DA uptake potency for cocaine and GBR12909, which is consistent with the single mutation of Y88F relative to WT hDAT 17 42 ; while cocaine is a competitive inhibitor for DAT and prefers to an outward-facing transport state, leading to a decrease in DA transport 43,44 . According to our computational modeling prediction, both hDAT residues Y88 and H547 are the critical binding sites on DAT for Tat 20 . Therefore, our findings demonstrate that (2019) 9:3843 | https://doi.org/10.1038/s41598-019-39872-1 www.nature.com/scientificreports www.nature.com/scientificreports/ through allosteric modulation, mutation of Y88 and H547 alters the inhibition potency for DA uptake and DAT binding site for substrate and DAT inhibitors 40 .
Our previous study has demonstrated that mutant H547A displays increased sensitivity to PKC-mediated downregulation of DAT function by PMA compared to WT hDAT 18 . Interestingly, the current study shows no difference in PMA-mediated reduction of DA uptake between Y88F/H547A and WT hDAT, indicating that Y88F/ H547A-induced increase in V max is independent of PKC activity. It has been shown that PKC-mediated phosphorylation of DAT downregulates DA uptake velocity [45][46][47] . Through allosteric modulatory effect, single mutation of H547 may alter the levels of PKC-induced phosphorylation of DAT activity, thereby causing the increased V max . However, double mutation of Y88 and H547 may attenuate the intermolecular structural interaction across PKC phosphorylation sites. Indeed, compared with WT hDAT, H547A, and Y88F/H547A enhances the V max values by 196% 18 and 96%, respectively. In addition to PKC-mediated regulation of DAT function, activation of palmitoyl acyltransferase, a group of enzymes that transfer palmityl group to -SH group on cysteine on a protein, enhances DA uptake via DAT palmitoylation 48 . Therefore, we will investigate whether Y88F/H547A-induced increased DA uptake is dependent on palmitoylation activity.
A major finding is that the double mutant Y88F/H547A completely attenuates the Tat-mediated inhibition of DAT reuptake function, which is consistent with our previous reports on Y88F-or H547A-induced attenuation of the Tat's effect 17,18 . According to our computational modeling prediction, either the Y88 or H547 residue could form hydrogen bonds with Tat residue K19 or R49 independently, suggesting that the double mutation of Y88 and H547 eliminated Tat-DAT binding. Hence, our pharmacological study further validates the computational prediction for the complementary hydrophobic interactions between Tat and hDAT. Furthermore, since we have demonstrated that Tat interacts with DAT allosterically 19,22 , the current findings also suggest that double mutation of Y88 and H547 attenuates Tat effect through a conformational mechanism. First, Y88F/H547A attenuates zinc-mediated increase in [ 3 H]WIN35,428 binding relative to WT hDAT. Second, Y88F/H547A augments basal DA efflux when compared to WT hDAT, which opposes our previous reports showing that neither Y88F nor H547A affected the basal efflux 17,18 . The elevated basal DA efflux in Y88F/H547A may be attributed to its elevated V max of DA uptake and DA accumulation. Taken together, Y88F/H547A, via an alteration of DAT conformational transitions, may influence the binding structure of hDAT-Tat complex, thereby disrupting Tat-induced inhibition of DA uptake. Studying the multiple mutation of DAT residues in Tat-DAT intermolecular interaction will greatly contribute to our ongoing proof of concept studies using novel allosteric modulators to establish their potential for therapeutic application in HAND.
AMPH, a substrate of DAT, competitively inhibits the reuptake of DA and elicits DAT-mediated DA efflux by reversal of the transporter 49 . Our computational modeling has demonstrated that the dynamic DA uptake process includes outward-open, outward-occluded, and inward-open states 21 , which can be reversed by DA substrate and DAT mutation [49][50][51] . The current results indicate that the potency for AMPH inhibition of [ 3 H]DA uptake was not altered in Y88F/H547A compared with WT hDAT, suggesting that the mutation of Tyr88 and His547 does not alter the interaction of AMPH with hDAT. However, AMPH-mediated DA efflux was reduced in H547A and Y88F/H547A, suggesting that mutations of Tyr88 and His547 residues alter the conformational equilibrium of the DAT transport transitions. Furthermore, AMPH decreases the V max of DA uptake in both WT hDAT and Y88F/ H547A, which is consistent with the previous report 51 , however, the magnitude of the AMPH-mediated reduction of the V max in Y88F/H547A was greater than that in WT hDAT. Interestingly, the AMPH-mediated reduction of the V max was accompanied by increased K m values in WT hDAT but not Y88F/H547A. Given that DAT-mediated DA transport is accompanied by reuptake of two Na + and Cl − ions, AMPH influx would transport Na + through DAT and enhance its levels at the inner face of the transporter 52 . Therefore, AMPH decreases DA transport but increases K m values in WT hDAT, whereas altered transporter conformation by Y88F/H547A may attenuate the increased K m . Taken together, the results along with attenuated zinc-regulated [ 3 H] WIN35,428 binding and increased basal DA efflux in Y88F/H547A further support that mutations of Tyr88 and His547 cause an alteration in conformational transitions, thereby disrupting the inhibitory effect of Tat on DAT reuptake function.
In conclusion, our data further demonstrate DAT Tyr88 and His547 as important residues for DAT function and the Tat-DAT interaction. Double mutation of these residues preserves single mutant H547A-mediated enhancement of DA uptake and diminishes Tat-induced inhibition of DA transport 18 . An allosteric modulatory mechanism may contribute to Y88F/H547A-mediated attenuation of Tat-induced inhibition of DA transport. Understanding the multiple mutation effects of the identified DAT residues might provide a basis for a novel approach for developing compounds to attenuate Tat binding to DAT by an allosteric mechanism in HIV-1-infected individuals.

Materials and Methods
Prediction of Tat-DAT binding model. We performed computational modeling and simulations to model the binding structure of hDAT with HIV-1 clade B type Tat based on the nuclear magnetic resonance (NMR) structures of Tat 53 and the constructed structure of hDAT-DA complex 21 . Protein docking and molecular dynamics simulations were employed to identify the conformation of hDAT-Tat complex. The energy-minimized complex structure used in this work were extracted from long-time equilibrated molecular dynamics simulation trajectories in our previous work 17,18,20,35 .

Construction of plasmids.
We selected the mutations of Y88, H547, Y88/H547 and Y88/K92/H547 based on predictions of computational modeling and simulations (Fig. 1) and our previous studies 17,18 . Single mutants, Y88F, K92M, and H547A, are expected to abolish hydrogen bonds between hDAT and Tat. Double (H547A/Y88F) and triple (H547A/Y88F/K92M) mutations were generated based on the wild type hDAT (WT hDAT) sequence (NCBI, cDNA clone MGC: 164608 IMAGE: 40146999) by site-directed mutagenesis. We then used the synthetic cDNA encoding hDAT subcloned into pcDNA3.1 + as a template to generate mutants using QuikChange ™ www.nature.com/scientificreports www.nature.com/scientificreports/ site-directed mutagenesis Kit (Agilent Tech, Santa Clara CA). We performed the DAT sequence for confirming the sequence of the mutant construct at University of South Carolina EnGenCore facility. Plasmids DNA were propagated and purified using a plasmid isolation kit (Qiagen, Valencia, CA, USA).
Cell culture and DNA transfection. In the current study, we used rat pheochromocytoma cells (PC12 cells, CRL-1721, American Type Culture Collection (ATCC), Manassas, VA), which were maintained in Dulbecco's modified eagle medium (DMEM, Life Technologies, Carlsbad, CA) supplemented with 15% horse serum, 2.5% bovine calf serum, 2 mM glutamine and antibiotics, 100 U/mL penicillin and 100 µg/mL streptomycin. Cells were cultured at 37 °C in a 5% CO 2 incubator. Once cells were growing to 100% confluence on plates, cells were seeded at a density of 1 × 10 5 cells/cm 2 in 24-well plates and transfected with WT hDAT or mutants by Lipofectamine 2000 (Life Technologies). Twenty-four hours after transfection, intact cells or cell suspensions were used for experiments. were applied to cells at room temperature for additional 8 min. The specific DAT-mediated DA uptake was calculated by subtracting non-specific uptake from total uptake (in the absence of nomifensine). The reaction was terminated by removal of solution from wells and quickly washed three times with ice cold KRH buffer. Cells were then lysed in 500 μL of 1% SDS for an hour and radioactivity was measured using a liquid scintillation counter (Tri-Carb 2900TR; PerkinElmer Life and Analytical Sciences, Waltham, MA).
To confirm whether Y88F/H547A-induced increase in V max was dependent on PKC regulatory mechanism, the V max of [ 3 H]DA uptake was measured in the presence or absence of 1 µM PMA, a PKC activator (Tocris, Bristol, UK) as the previous reports 29, 30 . Intact PC12 cells transfected with WT or Y88F/H547A were preincubated with or without PMA for 30 min at 37 °C followed by additional 8 min incubation with six concentrations of mixed [ 3 H]DA as described above. To determine the effect of amphetamine (AMPH) on Y88F/H547A-mediated transporter conformational transitions, V max or K m of [ 3 H]DA uptake in WT and Y88F/H547A was determined in the presence or absence of 1 µM AMPH (Sigma-Aldrich, St. Louis, MO) as reported previously 51 .
To determine whether these mutations alter the affinity of DAT substrate or inhibitors as well as Zn 2+ -regulation of DA transport, we performed the competitive inhibition of DA uptake in intact PC12 cells transfected with WT or mutant, which were preincubated with a series of final concentrations of DA (1 nM-1 mM), GBR12909 (1 nM-10 µM), cocaine (1 nM-1 mM), AMPH (1 nM-1 mM) or ZnCl 2 (1, 10, and 100 μM) at room temperature for 10 min followed by additional 8 min incubation with a fixed concentration of [ 3 H]DA (0.05 µM, final concentration).
In order to determine whether double mutation of Y88 and H547 disrupts Tat and DAT interaction, we performed the specific [ 3 H]DA uptake in PC12 cells transfected with WT or mutant in the presence or absence of Tat protein. In brief, cells were dissociated with trypsin/EDTA (0.25%/0.1%, 1 mL for one 10 cm dish) and resuspended in culture medium and incubated at room temperature for 10 min. The dissociated cells were collected by centrifugation at 400 × g for 5 min at 4 °C and washed once with phosphate-buffered saline followed by additional 5 min centrifugation (400 × g, 4 °C). Finally, the resulted cell pellets were resuspended in KRH assay buffer. The cell suspensions from WT hDAT or Y88F/H547A mutant were then preincubated with or without recombinant Tat 1-86 (rTat 1-86 , 140 nM, final concentration, Diatheva, Fano, Italy) at room temperature for 20 min followed by additional 8 min incubation with a mixed [ 3 H]DA uptake (0.05 µM, final concentration). Non-specific [ 3 H]DA uptake was determined in the presence of 10 µM nomifensine. The update reaction was terminated by immediate filtration through Whatman GF/B glass filters (presoaked with 1 mM pyrocatechol for 3 h) followed by three washes with 3 mL of ice-cold KRH buffer containing pyrocatechol using a Brandel cell harvester (model M-48; Brandel Inc., Gaithersburg, MD). Radioactivity was determined as described above.  WIN 35,428 was determined by adding in the presence of 30 µM cocaine (final concentration) and subtracted from total binding to obtain the specific binding. For competitive inhibition experiments, assays were performed in a final volume of 250 μL. Intact PC12 cells transfected with WT or mutated hDATs were incubated in buffer containing 25 μL of [ 3 H]WIN 35,428 (final concentration, 5 nM) and one of ten concentrations of DA (1 nM-100 μM), cocaine (1 nM-100 μM), GBR12909 (0.01 nM-1 μM) or ZnCl 2 (1, 10, 100 μM) on ice for 2 h. Assays were terminated by removing reaction reagents in each well followed by two washes with ice-cold assay buffer. Cells were lysed with 1% SDS for an hour and subjected to the radioactivity measurement as described above.
[ 3 H]DA efflux assay. Basal DA efflux was performed at room temperature as described previously [16][17][18] . Intact PC12 cells transfected with WT hDAT or its mutant were preloaded with 0.05 μM [ 3 H]DA for 20 min and then washed 3 times with KRH buffer prior to collecting fractional efflux samples. To obtain an estimate of the total (2019) 9:3843 | https://doi.org/10.1038/s41598-019-39872-1 www.nature.com/scientificreports www.nature.com/scientificreports/ amount of [ 3 H]DA in the cells at the zero time point, cells from a set of wells (four wells/sample) were lysed rapidly in 1% SDS after preloading with [ 3 H]DA. To collect factional efflux samples, buffer (500 μL) was added into a separate set of cell wells and transferred to scintillation vials after 1 min as an initial fractional efflux, and another 500 μl buffer was added to the same wells and collected after 10 min as second fractional efflux. Additional fractional efflux at 20, 30, 40, 50 min, respectively, was repeated under the same procedure. After last fractional efflux, cells were lysed and counted as total amount of [ 3 H]DA remaining in the cells from each well.
To further determine DAT mutation-mediated transporter confirmation transitions, basal efflux and AMPH-stimulated DA efflux in WT hDAT and its mutants were measured as reported previously 51 . PC12 cells transfected with WT hDAT and its mutants growing in 24-well plates were preloaded with [ 3 H]DA (50 nM, final concentration) for 20 min at room temperature. Subsequently, cells in all plate wells were washed with KRH for three times, cells were then incubated with KRH buffer containing AMPH (5-250 µM) or vehicle (KRH buffer alone, as baseline control) for 15 min at room temperature. After incubation, the buffer solution incubated with AMPH or vehicle was collected as AMPH-stimulated efflux or baseline efflux. Cells were lysed in 1% SDS and subject to count for obtaining the initial amount of DA (total DA) in the cells.

Data analysis.
In the current study, all data are presented as mean ± SEM with n as the number of independent experiments for each experiment group. Pharmacological parameter values (V max , K m , B max , and K d ) and IC 50 from [ 3 H]DA uptake and [ 3 H]WIN 35,428 binding were calculated by nonlinear regression analysis using a one-site model with variable slope. To determine the significant difference in the pharmacological parameters between unpaired sample groups, such as WT hDAT and its mutants, unpaired Student's t test was used for the statistical comparisons. Significant differences between sample groups were analyzed with separate ANOVAs followed by post-hoc, Bonferroni multiple comparison test as indicated in the results Section of each experiment. All statistical analyses were performed using IBM SPSS Statistics version 25, and differences were considered significant at p < 0.05.