Allosteric modulatory effects of SRI-20041 and SRI-30827 on cocaine and HIV-1 Tat protein binding to human dopamine transporter

Dopamine transporter (DAT) is the target of cocaine and HIV-1 transactivator of transcription (Tat) protein. Identifying allosteric modulatory molecules with potential attenuation of cocaine and Tat binding to DAT are of great scientific and clinical interest. We demonstrated that tyrosine 470 and 88 act as functional recognition residues in human DAT (hDAT) for Tat-induced inhibition of DA transport and transporter conformational transitions. Here we investigated the allosteric modulatory effects of two allosteric ligands, SRI-20041 and SRI-30827 on cocaine binding on wild type (WT) hDAT, Y470 H and ﻿Y﻿88 F mutants. Effect of SRI-30827 on Tat-induced inhibition of [3H]WIN35,428 binding was also determined. Compared to a competitive DAT inhibitor indatraline, both SRI-compounds displayed a similar decrease (30%) in IC50 for inhibition of [3H]DA uptake by cocaine in WT hDAT. The addition of SRI-20041 or SRI-30827 following cocaine slowed the dissociation rate of [3H]WIN35,428 binding in WT hDAT relative to cocaine alone. Moreover, Y470H and Y88F hDAT potentiate the inhibitory effect of cocaine on DA uptake and attenuate the effects of SRI-compounds on cocaine-mediated dissociation rate. SRI-30827 attenuated Tat-induced inhibition of [3H]WIN35,428 binding. These observations demonstrate that tyrosine 470 and 88 are critical for allosteric modulatory effects of SRI-compounds on the interaction of cocaine with hDAT.


Authors
Wei-Lun Sun, Pamela M. Quizon, Yaxia Yuan, Wei Zhang, Subramaniam Ananthan, Chang-Guo Zhan, and Jun Zhu This article is available at UKnowledge: https://uknowledge.uky.edu/mmbc_facpub/11 resulting in viral protein release, which has been implicated in the pathophysiology of HAND 20 . Cocaine abuse has been shown to increase the incidence of HAND and exacerbate the severity of HAND by enhancing viral replication [21][22][23][24][25][26][27] . Currently, there are no promising therapeutic approaches for cocaine addiction and HIV infection associated comorbidities 28 . Therefore, there is a pressing need to define the molecular mechanism(s) by which the impaired dopaminergic system by HIV-1 infection affects the progression of HAND in concurrent cocaine abusers.
The presynaptic dopamine transporter (DAT) plays an essential role in dopamine homeostasis and maintaining stable synaptic dopaminergic tone involved in attention, learning, memory 4,5 , and motivation 6,7 . Cocaine acts as a non-translocated inhibitor and exhibits non-selective binding to the DAT, serotonin transporter and norepinephrine transporter. However, the strong psychoactive behavioral responses and addictive effects of cocaine are mediated almost exclusively by its interaction with the DAT 29,30 . DAT is a primary target for cocaine binding, which has been shown to overlap DA uptake site 31 . In addition to competitive inhibitors and substrates of DAT, there is growing interest in allosteric modulation of DAT. Allosteric sites on human DAT (hDAT) may represent novel drug targets that display neutral cooperativity with the classical DA uptake site. There are a number of advantages in using allosteric modulators of DAT as preferred therapeutic agents over classic competitor of the DA uptake site with minimal effects on the basal DA transmission but decreasing the cocaine's action on DAT. For example, it has been shown that allosteric modulators of DAT such as the SRI-compounds act as partial antagonists of DA uptake without the full inhibitory profile that is typical of classic competitors of DAT [32][33][34] . In rat synaptosomes, SRI-compounds diminish cocaine's ability to inhibit DA uptake 35 , however, their effect on the interaction between cocaine and hDAT is still unknown. Further, it is uncertain whether the SRI-compounds suppressive effect on cocaine inhibition of DA uptake is mediated through their interaction with DAT, since these compounds also partially inhibit both serotonin and norepinephrine transporters 36,37 .
HIV-1 viral proteins are associated with the persistence of HIV-related neuropathology and subsequent neurocognitive deficits [38][39][40][41] . Among viral proteins, Tat protein plays a crucial role in the neurotoxicity and cognitive impairment evident in neuroAIDS 42,43 . DAT activity is strikingly reduced in HIV-1-infected cocaine-using patients, correlating with the severity of HIV-1 associated cognitive deficits 8,9 . We have demonstrated that Tat directly binds to DAT 44,45 . Exposure to Tat alone results in an inhibition of DA transport and promotes the internalization of DAT 44,46,47 . Interplay of Tat and cocaine augments synaptic DA levels and Tat release by inhibiting DAT activity 45,48 , which may contribute to the progression of HAND underlying the cognitive deficits in HIV-1 positive cocaine-using individuals 8,9 . Similarly, conditioned expression of Tat in the mouse brain further potentiates cocaine rewarding in vivo 49 . Together, these results suggest a synergistic effect of cocaine and Tat on DA transmission contributing to cognitive dysfunction and elevating the cocaine additive effects.
Our previous studies showed that Tat displays an allosteric modulatory effect on DAT function 35,45 . Through integrated computational modeling prediction and experimental validation, we explored structural regions of Tat-hDAT model to identify potential binding sites for Tat 50 . The structure of human DAT was homology modeled based on the X-ray crystal structure of drosophila DAT (dDAT) 51 in our previous studies 47,50,52 . The sequence identity between hDAT and dDAT is 46%, which is considered to be sufficient for constructing a satisfactory homology model 53,54 . We identified that DAT tyrosine 470 and 88 replaced by histidine (Y470H) or phenylalanine (Y88F) retained the normal surface DAT expression and inhibited Tat-induced inhibition of DA transport 44,47 . Further, mutating these two residues prevented zinc-induced conformational transporter transitions 44,47 , which may suggest that these two residues are critical for Tat allosteric modulation of DAT. In this study we first evaluated the effects of SRI-20041 and SRI-30827 (see Fig. 1), novel allosteric modulators of DAT 32 , on cocaine-mediated inhibitory effect on hDAT function. Further, we assessed whether the tyrosine 470 and 88 are critical for the allosteric modulation of hDAT by the SRI-compounds in modulating cocaine's interaction with DAT. Finally, we determined whether SRI-compounds attenuates Tat-induced inhibition of DAT binding sites. Understanding the allosteric sites of DAT may uncover new therapeutic avenues for treating both cocaine addiction and cocaine-accelerated HAND in HIV-1 positive individuals.

Effect of SRI-30827on Tat-induced inhibition of [ 3 H]WIN35,428 binding in WT hDAT.
We have demonstrated that Tat protein produces inhibitory effects on DAT function and expression. To determine whether the Tat-induced inhibition of DAT binding sites is mediated through an allosteric modulation manner, the ability of SRI-30827 to attenuate the inhibitory effect of Tat was determined (Fig. 4). Two-way ANOVA revealed that the main effect of Tat

Discussion
The current study reports the allosteric modulatory effects of SRI-compounds on cocaine and Tat binding to hDAT. This study is an extension of our previous work demonstrating that mutations of Y470 and Y88 of hDAT are critical for Tat-induced inhibition of DA transport and conformational transporter transitions 44,47,50 . There are three major findings from the current study. First, using indatraline to assess competitive DAT inhibition, we found that the SRI-compounds and indatraline produce a similar increase in the IC 50 value of inhibition of [ 3 H] DA uptake by cocaine in WT hDAT; however, the SRI-compounds-induced change in the IC 50 of cocaine is attenuated in Y470H and Y88F. Second, using SRI-compounds to assess allosteric modulation of DAT function, the addition of either SRI-20041 or SRI-30827 following the addition of cocaine significantly slowed the dissociation rate of [ 3 H]WIN35,428 binding compared to cocaine alone, which was totally and partially blocked in Y470H and Y88F, respectively. Third, SRI-30827 attenuates Tat-induced inhibition of [ 3 H]WIN35,827 binding in WT hDAT. The significance of the current observations is two-fold: 1) this is the first report on the key residues responsible for the allosteric modulatory effects on cocaine and Tat binding to hDAT, and 2) the Y470 and Y88 may be related to the allosteric modulatory effects for the SRI-compounds and the Tat binding to hDAT. These findings provide a novel mechanistic basis for developing compounds that specifically attenuate cocaine and Tat binding site(s) in hDAT to normalize DA transmission to physiological levels in HIV-infected cocaine-using patients.
HIV-1 Tat protein and cocaine synergistically disrupt normal physiological dopaminergic transmission by inhibiting DAT. Attenuating inhibitory effect of Tat and cocaine on DA transport is important for preventing the DAT-mediated dysfunction of DA system in HIV infected patients with cocaine abuse. Cocaine binding sites are overlapped with DA uptake sites 31 , which makes it impossible to generate a competitive inhibitor of cocaine binding that treats cocaine addiction without itself inhibiting DA uptake. However, if inhibition of DA uptake by cocaine is the result of an allosteric mechanism, it would be possible, at least in theory, to generate an allosteric modulator for treatment of cocaine addiction that might attenuate cocaine binding without affecting DA transport. In general, drugs that interact with DAT are typically classified into two categories: 1) inhibitors, such as     58,59 . Thus, characterization of the binding sites in DAT responsible for cocaine-induced conformational changes will provide a mechanistic basis to identify targets on the DAT for developing novel compounds that reduce cocaine's potency via their allosteric modulation of DA transport process. The present study shows that SRI-20041 and SRI-30827 allosterically modulate inhibition of [ 3 H]DA uptake by cocaine. Moreover, DAT dissociation assays show that the addition of SRI-20041 or SRI-30827 after cocaine significantly slows the dissociation rate. This is consistent with previous reports from our laboratory 35 and by others 32,36 , suggesting that these two SRI-compounds modulate cocaine binding to DAT through an allosteric modulation mechanism. Notably, the previous studies were conducted in rat striatal synaptosomes while the current study is the first to determine the allosteric modulatory effects of SRI-compounds in PC12 cells expressing hDAT. Given the important role of DAT in addictive effects of cocaine 29,30 , determining the effects of SRI-compounds in cocaine-mediated behavior change is an interesting topic for future investigations. The most intriguing observation is that mutants Y470H and Y88F do not appear to interact with cocaine but lead to an attenuation of allosteric modulation of SRI-compounds on cocaine-mediated inhibition of DA transport and cocaine-induced dissociation of [ 3 H]WIN35,428 binding. Our computational modeling predicts that both cocaine and Tat interact with hDAT in the outward-open conformational state, resulting in a synergistic inhibitory effect on DA transport 50 . Interestingly, Y470H and Y88F displayed a higher efficacy in inhibiting DA uptake and DAT binding by cocaine compared with WT hDAT 44,47 . In addition, the previous results show that Y470H-and Y88F-mutations altered zinc-induced regulation of DA uptake and WIN35,438 binding 44,47 . One possibility is that Y470H-and Y88F-mediated transporter conformational transitions may contribute to these changes in cocaine-mediated inhibition of DA uptake and dissociation in these mutants. Earlier studies have shown that Phenylalanine472 and Leucine475 that are in proximity of Y88 and Y470 are critical for the interaction of cocaine with DAT 31, 60 , indicating a complex interaction of cocaine and Tat with DAT. The complex interaction of cocaine and Tat may change the sensitivity to cocaine. This may contribute to the clinical and preclinical observations showing an acceleration of HAND in HIV-infected patients who abuse cocaine 8,9,13 and the neurocognitive/behavioral deficits observed in HIV-1 transgenic animals 49,61 . Therefore, these results suggest that Y470H and Y88F play a critical role in the synergistic effects of cocaine and Tat on DAT.
This study provides novel mechanistic insight that Y470 and Y88 residues may influence the allosteric modulation sites on hDAT for both SRI-20041 and SRI-30827. We have demonstrated that Tat-induced inhibition of DAT is mediated by allosteric binding site(s) on DAT, not the interaction with the DA uptake site 35,45,50 . Tat molecule binds to DAT through intermolecular electrostatic attractions and complementary hydrophobic interactions 50 . In particular, eliminating the cation-π interaction between Y470 in hDAT and M1 in Tat by mutating Y470 into a residue such as histidine or alanine would significantly weaken the binding between hDAT and Tat 47 . Further, in light of the Tat-DAT binding structure, mutation of Y88 to a residue without hydrogen-bonding capacity would also weaken the Tat-DAT binding 47 . These computational predictions are validated by experimental data showing that Tat-induced inhibition of DAT is attenuated in Y470H and Y88F 44,47 . Furthermore, Y470H but not Y470F attenuates Tat-induced inhibition on DAT because mutating Y470 to phenylalanine (Y470F) does not affect the cation-π interaction 47 . In the current study, we also found that SRI-compounds-induced decrease in cocaine IC 50 was attenuated in Y470H but not in Y470F, further supporting the important role of Y470 in the Tat-DAT interaction. By using a hDAT homology model to dock Tat into the transporter and MD simulations to probe the conformational state of hDAT bound to Tat, we found that Tat can only bind to the outward-open structure with favorable binding energies 50 . Therefore, we predicted that Tat binding would block the entry pathway of the dopamine substrate, thereby inhibiting dopamine clearance from the presynaptic cleft. Since the conformational changes in DA transport process involve conversions between the outward-and inward-open conformations 57 , we then conducted docking studies using homology models of hDAT at the different conformational states and found that the two SRI-compounds could only fit into the outward-open hDAT model (Fig. 1). Further, both SRI-20041 and SRI-30827 interact with DAT extracellular loop 6 (EL6) that contacts directly with Tat and can partially inhibit DAT uptake function with a similar chemical structure 36 . According to our previous modeling results, in the outward-open hDAT conformation, Y470 extends to the extracellular region where it interacts directly with Tat residues 50 . In the SRI-compound docked models, SRI-compounds may potentially influence the conformation of residues Tyr470 and Tyr88 with EL6 region, and thus likely modulate the binding of Tat on hDAT via an allosteric modulatory mechanism. Hence, although SRI-compounds may not interact directly with either Tyr470 or Tyr88 for competing with Tat binding, they can weaken the Tat DAT binding by changing the DAT conformation allosterically. This has been verified by our experimental data showing that SRI-30827 attenuates Tat-induced inhibition of DAT binding sites. Taken together, considering both Y470 and Y88 are associated with hDAT-Tat interactions, developing compounds directly targeting the specific binding sites on hDAT for Tat could be a viable approach for treatment of Tat-induced dopamine dysfunction. Alternatively, developing DAT-based allosteric modulators interacting with the specific residues that are structurally distinct from Tat binding sites would be another possible therapeutic approach.
In conclusion, we have identified that Y470 and Y88 may act as allosteric modulatory sites on hDAT responsible for the effects of cocaine binding displayed by the two allosteric ligands, SRI-20041 and SRI-30827. Given that mutation of these two residues (Y470H and Y88F) attenuates Tat-induced inhibition of DAT function 44,47 , our current study also demonstrates an allosteric modulatory effect of SRI-30827 on attenuation of Tat-induced inhibition of DAT binding sites. These findings presented herein raise the exciting possibility of potential therapeutic intervention for HIV infected patients with concurrent cocaine abuse. Proof of this concept could emerge from efforts directed toward discovery and development of candidate in vivo probe molecules with the desired allosteric modulation profiles coupled with favorable drug-like attributes. The effectiveness of an early intervention for HAND to preserve neurocognitive functions in HIV-infected individuals may ultimately depend on a treatment approach that combines compound(s) that specifically attenuate Tat binding site(s) in DAT with antiretroviral therapy, without affecting the normal function of DAT.

Methods
Materials. PC Fig. 1, molecular docking was performed to determine the structures of hDAT binding with SRI-compounds by using the AutoDock Vina 62 based on the constructed hDAT-Tat complex from our previous work 50 . Both the substrate-binding site and extracellular region of hDAT were defined as the search space for the docking. For each SRI-compound, 20 independent docking processes were performed in order to explore all possible binding conformations in such a large search space. The best protein-ligand binding conformation for each SRI-compound was selected according to the best binding score given by the AutoDock Vina. Further, the atomic charges for SRI-compound molecule were determined as restrained electrostatic potential (RESP)-fitted charges based on the first-principles electronic structure calculations at the B3LYP/6-31 G* level by using the Gaussian03 program 63 . For each SRI-compound, the selected binding conformation was combined with the constructed hDAT-Tat structure with tLeap of the AMBER12 package 64 . Then the SRI-hDAT-Tat complex was optimized with 100,000 steps of the deepest decent energy-minimization and 100,000 steps of the conjugate gradient energy-minimization by using the Sander module of the AMBER 12 package 64 .

Molecular Modeling. As shown in
Construction of plasmids. Point mutations of hDAT at Y470 (tyrosine to histidine, Y470H-hDAT; tyrosine to phenylalanine, Y470F-hDAT) and Y88 (tyrosine to phenylalanine, Y88F-hDAT) were selected based on the structural and mechanistic insights obtained from computational modeling and simulation and previous studies 47,50 . The Y470H-hDAT is expected to eliminate both hydrogen bond and cation-π interactions in the native structure of the transporter. Both Y470F-hDAT and Y88F-hDAT are expected to preclude the formation of hydrogen bond. All mutations were generated according to WT-hDAT sequence (NCBI, cDNA clone MGC: 164608, IMAGE: 40146999) by QuikChange TM site-directed mutagenesis kit (Agilent Tech, Santa Clara, CA) with synthetic cDNA encoding hDAT subcloned into pcDNA3.1 + (a gift provided by Dr. Haley E Melikian, University of Massachusetts) as the template. The sequences of mutated constructs were further confirmed through DNA Scientific RepoRts | 7: 3694 | DOI:10.1038/s41598-017-03771-0 sequencing at University of South Carolina EnGenCore facility. Plasmid DNA were propagated and purified using a plasmid isolation kit (Qiagen, Valencia, CA).
Cell culture and DNA transfection. PC12 cells were maintained in the incubator (37 °C, 5% CO 2 ) with DMEM medium supplemented with horse serum (15%) and bovine calf serum (2.5%), glutamine (2 mM), and penicillin-streptomycin (100 U/ml). Chinese hamster ovary cells (CHO cells, ATCC, CRL-61) were maintained in F12 medium with 10% fetal bovine serum and antibiotics (100 U/mL penicillin and 100 µg/mL streptomycin). For hDATs transfection, cells were seeded into poly-D-lysine-coated 24 well plates at a density of 1 × 10 5 cells/well. After 24 h, cells were transfected with WT or mutant DAT plasmids using Lipofectamine 2000 (Life Tech, Carlsbad, CA) based on the manufacturer's instruction. Cells were used for the experiments after 24 h of transfection.

Competitive inhibition [ 3 H]DA uptake assay.
Considering that PC12 cells endogenously expressing norepinephrine transporter (NET) and DA is transported by DAT and NET, in a pilot study, we first performed kinetic analysis of [ 3 H] DA uptake in PC12 cells with no transfection in the presence of desipramine (1 µM, final concentration) to prevent DA uptake into norepinephrine containing nerve terminals based on a previously published method 65 . Results showed that no specific [ 3 H] DA uptake was found with a range of DA concentrations (1 nM-5 µM, data not shown).  on ice for 2 hr in duplicate. After incubation, the intact cells in each well were washed twice with ice-cold assay buffer, lysed by 1% SDS for 1 hr and subjected to liquid scintillation spectrometry measurement. For the binding dissociation assay, experiments were conducted in a final volume of 500 µl according to our previous publication 35 . The intact cells were incubated with a fixed concentration of [ 3 H]WIN 35,428 (5 nM) on ice for 2 hr followed by a wash of ice-cold assay buffer. Non-specific [ 3 H]WIN 35,428 binding was determined by addition of β-CFT naphthalenedisulfonate monohydrate (1 µM). In condition 1 (cocaine only), at the zero time point, the binding dissociation was initiated by the application of a single concentration of cocaine (1 µM). In condition 2 (cocaine + SRI-compounds), 10 min after the addition of cocaine, SRI-20041 (10 µM) or SRI-30827 (50 nM) was then added to minimize any [ 3 H]WIN35,428 re-association. Ten, 20, 30, 40, and 60 min later, duplicate wells of cells were washed twice with assay buffer, lysed, and counted by liquid scintillation spectrometry. For data analyses, 0 min (no drugs treatment) and 10 min after the application of cocaine were set as 100% to normalize samples at each time point in condition 1 and 2, respectively, based on previous publications 32, 35 .
Our previous study shows that Tat inhibits [ 3 H]WIN35,428 binding, which is consistent with the ability of Tat inhibiting DA uptake 45 . To determine the effect of SRI-30827 on Tat-induced inhibition of DAT binding sites, the specific [ 3 H]WIN35,824 binding (5 nM, final concentration) in WT hDAT was measured in the presence or absence of SRI-30827 (50 nM, final concentration) or recombinant Tat 1-86 (40 nM, final concentration, ImmunoDX, Woburn, MA) on ice for 2 h. This concentration of SRI-30827 was chosen because our pilot study showing SRI-30827 at this concentration did not alter basal [ 3 H]WIN35,824 binding and also this concentration was used for the dissociation assay. Nonspecific [ 3 H]WIN35,824 binding was determined in the presence of 30 µM cocaine.
Data analyses and Statistics. Descriptive statistics and graphical analyses were used as appropriate.
Results are presented as mean ± SEM, and n represents the number of independent experiments for each experiment group. IC 50 values for cocaine inhibition of [ 3 H]DA uptake and [ 3 H]WIN35,428 binding were calculated from inhibition curves by nonlinear regression analysis with a one-site model of variable slope. For the dissociation of [ 3 H]WIN35,428 binding induced by cocaine, the dissociation rate (K −1 ) was determined by the specific [ 3 H]WIN35,428 binding through non-nonlinear regression analysis using a single component dissociation model. Both kinetic parameters (cocaine IC 50 and K −1 ) were calculated by Garphad Prism version 5.0 (GarphPad Software Inc., San Diego, CA). For experiments involving comparisons between unpaired samples, separate ANOVAs followed by appropriate post hoc tests [Fisher's least significant difference (LSD) or Student's unpaired t test] were used. All statistical analyses were performed using IBM SPSS Statistics version 24. p < 0.05 was the minimum criterion for statistical significance.