DBP1: Dynamics of Neurotransmitter Transporters: Molecular and Cellular Interactions

Back to table

A. Collaborating Investigators: Susan G Amara,1 Gonzalo Torres,2 Delany Torres-Salazar,1 Jennie Garcia-Olivares,1 Ivet Bahar,3 Mary Cheng 3 Bing Liu,3 James Faeder,3 Terry Sejnowski,4 Tom Bartol4

B. Institutions: 1NIH, 2University of Florida, 3Pitt and 4Salk

C. Funding Status of Project: 1ZIAMH002946-03 (Amara); R01 DA038598 (Torres) 9/15/14 - 6/30/19

D. Driving relationship between DBP1 and TR&Ds. Sodium-coupled neurotransmitter (NT) transporters regulate neurosignaling in the CNS and prevent neurotoxicity by clearing excess NT from the synapse. Significant studies have been made in recent years for elucidating their mechanisms of function, including the contributions of TR&D1 and DBP1 during the past term to characterizing (i) the structure-encoded dynamics of excitatory amino acid transporters (EAAT1-3) and dopamine transporter (DAT) or its structural homolog LeuT1-5 and (ii) the molecular basis of their cellular function/dysfunction,6-11 Yet, several aspects of NT transport remain unknown. First, many transporters also act as anion (e.g. chloride, Cl-) channel. Anion conductance has been suggested to promote electrogenic Glu- uptake12 and serve as a sensor for regulating the release of additional Glu-.13,14 The Amara lab and others15-18 found that Cl- channeling is structurally coupled to Glu- transport, but the molecular mechanism of coupling is unclear. Second, NT transporters also allow for the efflux (or reverse transport) of NT, modulated by regulatory proteins such as G protein β-γ subunits (Gbg)6 and CaMKII,19-21 addictive drugs such as cocaine and amphetamine (AMPH), or lipids.22-24 How the transporter structure adapts to enable NT efflux remains to be understood. Third, recent studies by the Amara lab highlight the key role of intracellular (IC) AMPH in modulating the function and internalization of transporters;25,26 it is not clear how protein-protein interactions (PPIs) stimulated by AMPH alter or impair neurosignaling. This DBP aims at shedding light on these fundamental questions. The first two require molecular level computations, which will be conducted in TR&D1 aim 1, driven by the molecular biology, site-directed mutagenesis, substituted Cys accessibility (SCAM), biochemical and electrophysiology experiments in the Amara and Torres labs. The third topic, will drive TR&D3 aim 2.3, and will be pursued in in collaboration with TR&D1 (aims 1 and 3) and TR&D2 (aim 2.3). Therefore, DBP1 will drive TR&D1-3, and we also anticipate a synergistic collaboration with DBP3, as described below in the Methods and Procedure below in the context of the 3 aims of the DBP1.


E. Innovation: We will develop (i) an advanced protocol that will permit users to combine ENMs, aMD, and co-MD27,28 to explore substrate/ion permeation paths; (ii) automated methods for evaluating disulfide trapping and solvent/reagent accessibility experiments; (iii) the first comprehensive quantitative network model for AMPH-stimulated signaling in DA neurons. Combined experimental and computational results will provide a mechanistic understanding of the orchestration of IC AMPH-triggered PPIs and an in silico platform for developing new hypotheses for controlling AMPH actions or regulating neurotransmission.

F. Methods and Procedure: Aim 1. Test the hypothesis that Cl-/anion channeling by EAAT1 involves a structural change into an intermediate state and identify the location of Cl- permeation pore and key residues lining the pore. We presented in TR&D1 Aim 1 methods and initial results (see Fig III.4 therein) on the computationally predicted pore for Cl- channeling, and on critical residues, e.g. GltPh F50 and V51, along the channel. Using the SCAM method, the Amara lab heterologously expressed in oocytes, single Cys substitutions at EAAT1 residues predicted by the Bahar lab (M89C, counterpart of Gltph V51) to be within the anion permeation pathway. In at least 3 of the substitutions we observed an approximately 50% reduction in the macroscopic current amplitude after application of MTSES, a negatively charged sulfhydryl-reactive reagent. We observed comparable results with MTSACE, a similar, but non-charged reagent, ruling out that the observed current reductions were caused by electrostatic interactions. Moreover, our results consistently showed that substrate translocation is not altered by these modifications (Fig VII.1).Further experiments and computations with homology-modeled EAAT1 are in progress for consolidating these promising results. Computations also suggest a switch to an open intermediate mediated EAAT1 R388 and E377. Experiments with the double mutants R388E-E377R and R388E-D276R will test the hypothesis that these compensating mutations might restore WT behavior.


Fig VII.1. Application of 10 mM MTSES reduced anion current amplitudes but not substrate transport. (A) Time dependence of MTSES modification of anion (NO3-) macroscopic current amplitudes measured at +60 mV in oocytes expressing the mutant M89C (τ= 5.2 ± 0.2 seconds; n=5). Black bar indicates the duration of the application. (B) Current-voltage relationship of the same cells represented in A before (red circles) and after (cyan triangles) application of MTSES, and after application of 1mM DTT (green diamonds) following the application of MTSES. (C) Bar representation of the current amplitudes at +60 mV from B. (D) Current-voltage relationship representing transport currents before (red diamonds) and after (cyan diamonds) the application of MTSES (currents were measured in chloride-based solution; n=5). (E) Bar representation of the currents at -120 mV from D. (F) radiolabeled Glu- uptake before and after application of MTSES remains unchanged (n = 34).

Aim 2. Test the hypotheses that G-βγ binding to DAT drives an allosteric conformational change in DAT, which in turn, promotes DA efflux. We generated a first structural model for DAT-Gβγ complex (see Fig III.5 in TR&D1 section) and identified two interfacial salt bridges between Gβγ and hDAT C-terminus (R588 and R610). DA efflux assays by Torres using the mutants R588A and R610A showed preliminary results in support of our model. Our simulations of hDAT-Gbg complex in explicit water and lipid revealed a water channel (Fig III.5B, manuscript in preparation), which is to our knowledge the first such observation in hDAT. This may be indicative of a possible anion channel, or DA efflux path. Notably, highly conserved D79 and D421 line the DA translocation path;29 whereas E117 and E491 line this new water channel. We will establish the role of this channel by iterative, coordinated computations including metadynamics 30 to quantify selective ion permeation (see e.g.31) and experiments (mutagenesis, immunoprecipitations and functional assays).


Aim 3. Model and analyze the dynamics of the PPI network that mediates AMPH-stimulated DAT internalization and other signaling events in DA neurons. New data from the Amara lab,32 along with those from the literature and models determined in TR&D1, will be used to build a comprehensible BioNetGen (TR&D3) model for Rho- and PKA-dependent AMPH actions, which will be iteratively tested and validated against time-resolved concentration and activity data to be supplied by the Amara lab. TR&D3 subaim 2.3 present the methods together with initial results in support of the adopted model. The further calibrated and validated model and parameters will be used in MCell simulation of DA neurons (TR&D2) in collaboration with DBP3 (see also Fig I.10 in the Overview) .


Copyright © 2018 National Center for Multiscale Modeling of Biological Systems. All Rights Reserved.