An Intracellular Interaction Network RegulatesConformational Transitions in the Dopamine Transporter*□S

Received for publication, January 18, 2008, and in revised form, April 2, 2008 Published, JBC Papers in Press, April 21, 2008, DOI 10.1074/jbc.M800475200

Julie Kniazeff‡1,2, Lei Shi§1, Claus J. Loland‡, Jonathan A. Javitch¶, Harel Weinstein§3, and Ulrik Gether‡4

From the ‡Molecular Neuropharmacology Group and Center for Pharmacogenomics, Department of Neuroscience andPharmacology, The Panum Institute, University of Copenhagen, DK-2200 Copenhagen, Denmark, the §Department of Physiologyand Biophysics and the HRH Prince Alwaleed Bin Talal Bin Abdulaziz Alsaud Institute for Computational Biomedicine, Weill MedicalCollege of Cornell University, New York, New York 10021, and the ¶Center for Molecular Recognition and Departments ofPsychiatry and Pharmacology, Columbia University College of Physicians and Surgeons, New York, New York 10032

Neurotransmitter:sodium symporters (NSS)1 mediate sodium-dependent reuptake of neurotransmitters from the synapticcleft and are targets for many psychoactive drugs. The crystalstructure of the prokaryotic NSS protein, LeuT, was recentlysolved at high resolution; however, the mechanistic details ofregulation of the permeation pathway in this class of proteinsremain unknown. Here we combine computational modelingand experimental probing in the dopamine transporter (DAT)to demonstrate the functional importance of a conserved intra-cellular interaction network. Our data suggest that a salt bridgebetween Arg-60 in the N terminus close to the cytoplasmic endof transmembrane segment (TM) 1 andAsp-436 at the cytoplas-mic end of TM8 is stabilized by a cation-� interaction betweenArg-60 and Tyr-335 at the cytoplasmic end of TM6. Computa-tional probing illustrates how the interactions may determinethe flexibility of the permeation pathway, and mutagenesiswithin the network and results from assays of transport, as wellas the state-dependent accessibility of a substituted cysteine inTM3, support the role of this network in regulating accessbetween the substrate binding site and the intracellular milieu.The mechanism that emerges from these findings may beunique to the NSS family, where the local disruption of ionicinteractions modulates the transition of the transporterbetween the outward- and inward-facing conformations.

Upon neuronal stimulation, neurotransmitters are releasedinto the synaptic cleft where they activate both pre- andpostsynaptic receptors. The duration of action of the transmit-ters is tightly controlled by integral membrane transport pro-

teins situated in the presynaptic nerve terminal or on the sur-face of surrounding glial cells where they mediate rapidsequestering of the transmitters from the synaptic cleft. Neuro-transmitter:sodium symporters (NSS)5 (also called Na�/Cl�-dependent transporters) constitute the major class of thesetransport proteins and include the transporters for dopamine(DA), serotonin, norepinephrine, glycine, and �-aminobutyricacid (GABA) (1–4). NSS proteins operate by coupling trans-port of Na� down its concentration gradient with “uphill”transport of substrate. Moreover, NSS proteins are character-ized by co-transport of Cl� (5). Transporters in this family havereceived particular attention as targets for many drugs includ-ing antidepressants, antiepileptics, and psychostimulants, suchas cocaine and amphetamines (1–4).Little is known about the structural dynamics that underlie

the function of NSS proteins. Presumably, the transporters fol-low an alternating access model in which the binding site isalternately exposed to the extracellular (“outward facing” con-formation) and intracellular environments (“inward facing”conformation) (6, 7). It is envisioned that binding of substratetogether with sodium and chloride to the outward facing con-formation elicits a conformational change that shifts the trans-porter to the inward facing conformation, allowing release ofsubstrate and co-transported ions to the intracellular environ-ment. Notably, such a mechanism implies the existence of two“gates,” one external and one internal, capable of occludingaccess to the substrate binding site from the extracellular orintracellular environments, respectively.The first insight into the three-dimensional structure of this

class of proteins was achieved recently by crystallization of theprokaryotic NSSmember, LeuT (8–10). The structure revealeda conformation likely representing an intermediate betweenthe “outward” and “inward” facing conformations (8). At thecytoplasmic side, the structure suggested the existence of atight network of interactions thatmight serve as an intracellulargate (Fig. 1). Specifically, an ionic interaction was foundbetween Arg-51.26 in the N terminus close to the cytoplasmicend of transmembrane segment (TM) 1 and Asp-3698.74 at thecytoplasmic end of TM8 (generic numbers of residues in super-

* This work was supported, in whole or in part, by National Institutes of HealthGrant P01 DA 12408. This work was also supported by the Danish HealthScience Research Council, the Lundbeck Foundation, the Novo NordiskFoundation, and the Maersk Foundation. The costs of publication of thisarticle were defrayed in part by the payment of page charges. This articlemust therefore be hereby marked “advertisement” in accordance with 18U.S.C. Section 1734 solely to indicate this fact.

□S The on-line version of this article (available at http://www.jbc.org) containssupplemental Figs. S1 and S2 and Movies I and II.

1 Both authors contributed equally to this work.2 Recipient of a European Molecular Biology Organization long term

fellowship.3 To whom correspondence may be addressed: Rm. E-509, 1300 York Ave.,

New York, NY 10065. Tel.: 212-746-6358, Fax: 212-746-8690; E-mail:haw2002@med.cornell.edu.

4 To whom correspondence may be addressed. Tel.: 45-3532-7548; Fax:45-3532-7610; E-mail: gether@sund.ku.dk.

5 The abbreviations used are: NSS, neurotransmitter:sodium symporters; DA,dopamine; DAT, dopamine transporter; WT, wild type; TM, transmembranesegment; IL, intracellular loop; MTSET, [2-(trimethylammonium)ethyl]me-thane thiosulfonate; GABA, �-aminobutyric acid; HA, hemagglutinin; MD,molecular dynamics; ELISA, enzyme-linked immunosorbent assay.

THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 283, NO. 25, pp. 17691–17701, June 20, 2008© 2008 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in the U.S.A.

JUNE 20, 2008 • VOLUME 283 • NUMBER 25 JOURNAL OF BIOLOGICAL CHEMISTRY 17691

script, see “Experimental Procedures”). This interaction is oneof three that are mutually stabilized, including the cation-�interaction between Arg-5 and Tyr-2686.68, and the hydrogenbonding between Arg-5 and Ser2676.67 (Fig. 1 and Ref. 8). Fur-thermore, the side chain of Tyr-268 also interacts with Gln-3618.66 and Ile-1874.62, and altogether these residues form thenetwork illustrated in Fig. 1, D and E. The residues in this net-work are all highly conserved among NSS proteins and thuslikely of critical importance for transporter function (Fig. 1 andRef. 11). In agreement, prior to the availability of the LeuTstructure we had mutated in DAT the tyrosine (Tyr-335 inDAT) and aspartic acid (Asp-436 in DAT), and obtained evi-dence for a pivotal role of these residues (12, 13); however, wewere at that time unable to interpret the data in a relevantstructural context.Here, in the context of the recent LeuT structure and a DAT

homology model, we perform a detailed analysis of the con-served network in DAT. The experimental work is focused onthe putative salt bridge between Arg-60 (Arg-5 in LeuT) andAsp-436 (Asp-369 in LeuT). From a series of single and doublemutations, we show evidence that the salt bridge is indeed pres-ent also in DAT. Moreover, analysis of dynamic models andcomputational normal mode analyses of the protein suggeststogether with experimental efforts that this interaction net-work is critical for stabilizing the outward facing conformationof the transporter and for regulating conformational transitionsin the translocation cycle.

EXPERIMENTAL PROCEDURES

Materials—All chemicals were from Sigma unless statedotherwise.Indexing of Residues—A generic numbering scheme for

amino acid residues inNSS proteins has been proposed to facil-itate direct comparison of positions between the individualmembers of the family (11, 14). According to this scheme, themost conserved residue in each transmembrane segment hasbeen given the number 50, and each residue is numberedaccording to its position relative to this conserved residue. Forexample, 1.55 indicates a residue inTM1 five residues carboxyl-terminal to themost conserved residue in thisTM(Trp1.50). ForDAT, themost conserved residues in each transmembrane seg-ment is as follows (generic number being indicated in super-script): TM1, Trp-841.50; TM2, Pro-1122.50; TM3, Tyr-1563.50;TM4, Cys-2434.50; TM5, Pro-2735.50; TM6, Gln-3176.50; TM7,Ser-3667.50; TM8, Phe-4128.50; TM9, Phe-4579.50; TM10, Phe-47810.50; TM11, Pro-52911.50; and TM12, Pro-57312.50. Com-paredwith the generic numbering schemedefined inRef. 11,weextend the TM index into the immediate conserved loopregions, specifically TM1 to theNH2 terminus, TM6 to IL3, andTM8 to IL4.Site-directed Mutagenesis—Synthetic cDNAs encoding the

human DAT (synDAT) and E2C (a DAT construct whereCys-90 and Cys-306 are mutated to alanines) were subclonedinto pcDNA3 (Invitrogen) (13). All mutations were generatedby the QuikChangeTM method (adapted from Stratagene, LaJolla, CA). All mutations were confirmed by DNA sequencing.

Cell Culture and Expression—COS7 cells were grown asdescribed and transiently transfected with the indicated con-structs using the calcium phosphate precipitationmethod (13).[3H]DA Uptake Measurements—Uptake assays were per-

formed as described (13) using 2,5,6-[3H]DA (9–13 Ci/mmol)(Amersham Biosciences). Briefly, transfected COS7 cells wereplated in either 24-well dishes (105 cells/well) or 12-well dishes(3 � 105 cells/well) coated with polyornithine (Sigma). Theuptake assays were carried out 2 days after transfection for 5min at room temperature in uptake buffer (25 mM HEPES, 130mM NaCl, 5.4 mM KCl, 1.2 mM CaCl2, 1.2 mM MgSO4, 1 mML-ascorbic acid, 5 mM D-glucose, and 1 �M of the catechol-O-methyltransferase inhibitor Ro 41-0960 (Sigma), pH 7.4). Theindicated non-labeled compounds were added to the cells priorto initiation of uptake by addition of 40 nM [3H]DA. Nonspe-cific uptake was determined with 1mMDA. Note that the Zn2�

experiments were done in constructs without the HA tagbecause introduction of the HA tag mutates His-193, which isone of the key coordinates in DAT Zn2� binding sites (15).[3H]DA Uptake Assay with MTSET Preincubation—Two

days after transfection, the cells (3 � 104/well in a 12-well dish)were washed once in 500 �l of uptake buffer. The cells weresubsequently incubatedwith 0.5mMMTSET ([2-(trimethylam-monium)ethyl]methane thiosulfonate) (Toronto ResearchChemicals, Toronto, Canada) at room temperature for 10 min.The stock MTSET solution in H2O was freshly prepared anddiluted 10-fold directly into a total volume of 500 �l of uptakebuffer. After incubation, the cells were washed twice in 500 �lof uptake buffer before initiation of [3H]DA uptake performedas described above. The effect of Zn2� on MTSET reactivitywas investigated by addition of 10 �M of a ZnCl2 solution justbefore adding MTSET.ELISA for Quantification of Cell Surface Expression—Two

days after transfection, cells were washed twice with phos-phate-buffered saline and fixed in 4% paraformaldehyde. After30 min blocking of the unspecific site with phosphate-bufferedsaline supplemented with 5% fetal calf serum, anti-HA anti-body, coupled to the horseradish peroxidase (80 milliunits/ml,clone 3F10, Roche, Basel, Switzerland), was applied for 30 minat room temperature in the same buffer. After intense washeswith phosphate-buffered saline, the antibody was detected andquantified instantaneously by chemiluminescence using Super-signal ELISA femtomaximum sensitivity substrate (Pierce) anda Wallac Victor2 luminescence counter (PerkinElmer Life Sci-ence). All experiments were performed at least in triplicate.Data Calculations and Statistical Analysis—Uptake data

were analyzed by nonlinear regression analysis using Prism4.02fromGraphPad Software, San Diego, CA. The IC50 values usedin the estimation of Km were calculated from means of pIC50values and S.E. interval from pIC50 � S.E (16). The Ki valueswere calculated from the IC50 values using the equation,

Ki � IC50/�1 � �L � Km�� (Eq. 1)

in which L � concentration of [3H]DA. All values in the figuresare provided as mean � S.E. For comparisons betweens twogroups, t test (two-tailed) was performed.LeuT Model Construction and Normal Mode Analysis—The

construction and equilibration of a full-length LeuTmodel in a

An Intracellular Interaction Network in NSS Proteins

17692 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 283 • NUMBER 25 • JUNE 20, 2008

1-palmitoyl-2-oleol-sn-glycero-3-phosphocholine lipid bilayeris described in detail elsewhere.6 Briefly, the simulation systemswith the LeuTmolecule immersed in an explicit representationof the water/lipid bilayer/water environment were constructedwith VMD (17) and equilibrated with nanoscale moleculardynamics (18), following a procedure modified from a recentdescription (19). The transporter is imbedded in a membranepatch of 204 1-palmitoyl-2-oleol-sn-glycero-3-phosphocholinemolecules, 101 on the periplasmic side and 103 on the cytoplas-mic side. The entire system includes around 78,000 atoms, witha final dimension around 87 � 87 � 98 Å3. A Tyt1 homologymodel was constructed with Modeler (20) based on the LeuTstructure, and equilibrated in the lipid bilayer as describedabove for LeuT. The conformation shown in supplemental Fig.S1 was obtained at the end of 12 ns of free equilibration.Themutants were constructed with PyMOL (DeLano Scien-

tific, Palo Alto, CA) using its backbone-dependent rotamerlibrary (21). Side chain rotamers were chosen to avoid stericclashes with the rest of the transporter model without changesin the backbone. The resulting models are R5A and Y268A ofLeuT, which are aligned with the corresponding R60A andY335A mutations in DAT. We also built R193A of LeuT, aresidue facing lipid, as a control. The normal mode analysiswithin the elastic network model (16) was carried out on theElNemo server (22).

RESULTS

A Conserved Intracellular Interaction Network—A recentcomprehensive sequence alignment (11) highlighted a markedconservation in theN terminus/TM1, intracellular loop 3 (IL3),and TM8 of NSS proteins (relevant parts shown in Fig. 1a). Theresidues include Arg1.26, Ser6.67, Tyr6.68, and Asp8.74 that areconserved among most NSS members (except in Tyt1, seebelow). In the LeuT structure (8), these residues (Arg-51.26, Ser-2676.67, Tyr-2686.68, and Asp-3698.74) together with Ile-1874.62and Gln-3618.66 form a local network of interactions, intercon-necting the N terminus/TM1 to IL3 and TM8 (Fig. 1, c and d).Notably, this network could coordinate the opening and closingof what can be considered an intracellular “gate” (8). The cor-responding network in a LeuT-based DATmodel involves res-idues Arg-601.26, Val-2594.62, Ser-3346.67, Tyr-3356.68, andAsp-4368.74 (Fig. 1, b and e). In thismodel, the Tyr-3356.68 side chainis also stabilized by the carbonyl groupofGlu-4288.66 that alignswith Gln-3618.66 in LeuT, but is conserved as a glutamate inmost other NSS proteins (Fig. 1, b, d, and e).

In the interaction network, an ionic interaction between thearginine (Arg-5 in LeuT, Arg-60 in DAT) and aspartate (Asp-369 in LeuT, Asp-436 inDAT) is of essential importance (Fig. 1,d and e). The salt bridge is complemented by a cation-� inter-action of the arginine with the tyrosine (Tyr-268 in LeuT, Tyr-335 in DAT) that requires a specific dihedral angle preferencefor the arginine side chain and brings IL3 into the network ofinteractions with the N terminus and TM8. Of note, a similarcation-� interaction is found in the viral protein VP39 (ProteinData Bank code 1V39) (23) where the interaction between argi-

nine and tyrosine generates mutual stabilization of the interac-tion between arginine and aspartate (24).In a comparative molecular dynamics (MD) study for the

known structure of LeuT, we attempted to estimate the stabi-lizing effect of the cation-� interaction by examining theimpact of mutating the tyrosine to alanine (Y268A). In the MDtrajectories of wild type (WT), Arg-5maintained its interactionwith Ser-267 and Asp-369 (Fig. 2a, top panel), whereas inY268A, Arg-5 interacted with either Ser-267 or Asp-369, butnot with both, and the entire local network was more mobile(Fig. 2a, bottom panel). This suggests that the cation-� interac-tion between Tyr-268 and Arg-5 positions Arg-5 in the inter-action network to enable its effective interaction with bothSer-267 and Asp-369. In addition, the hydrogen bond betweenTyr-268 and Gln-361 within the intracellular network couldcontribute to stabilizing TM8 relative to the N terminus. Notealso that Y268A loses its connection to Gln-361, a residuelocated only one turn below the substrate binding site, whereasAla-358 and Ile-359 are in direct contact with substrate. Alto-gether, this in silico analysis supports the hypothesis that theregion is important for substrate release, as destabilization ofthe cation-� and salt bridge interactions appeared to changethe structure and flexibility of the local environment surround-ing the substrate binding site.The Role of Arg-60 in the Corresponding Interaction Network

in DAT—Previously, we showed in DAT that mutation of Tyr-335 to alanine (Y335A) dramatically reduced transport capacityand lowered the Km value for [3H]dopamine ([3H]DA) trans-port (12). The mutation also changed the effect of Zn2� at theendogenous high-affinity Zn2� binding site in DAT; whereasbinding of Zn2� to this site causes non-competitive inhibitionof transport in the WT DAT, Zn2� stimulates [3H]DA trans-port in Y335A (12, 15, 25). Because Zn2� is likely to stabilize thetransporter in its outward facing conformation (13), this phe-notype was proposed to result from a change in the conforma-tional equilibrium in Y335A toward the inward facing confor-mation; hence, Zn2� stimulated transport by partiallynormalizing the equilibrium between the inward and outwardfacing conformations (13).Because the tyrosine (Tyr-335) appears to stabilize the Arg-

60/Asp-436 salt bridge via a cation-� interaction, we reasonedthat disruption of the salt bridge would produce a similar phe-notype as that of Y335A. Accordingly, we mutated Arg-60 toalanine, and for comparison we analyzed in parallel the previ-ously described D436A (13). The mutations were made both intheWTDATbackground and in amodifiedDATcontaining anHA antibody tag inserted into the second extracellular loop 2.In agreement with previously published results (26), this tag didnot affect uptake properties of the transporter, i.e. Km andVmaxfor [3H]DA uptake in HA-WTwas 1.61 �M (S.E. interval 1.34–1.93) and 4310 � 736 fmol/min/105 cells (n � 8), respectively,versus 1.14 �M (S.E. interval 1.04–1.24) and 3547 � 177 fmol/min/105 cells (n� 3) for non-taggedWTDAT. Insertion of theextracellular HA tag permitted quantification of DAT surfaceexpression by ELISA; however, introduction of the HA tagremoved one of the Zn2� coordinating residues in the endoge-nous Zn2� binding site (His-193) (15, 25) making HA-WTinsensitive to Zn2� modification. Consequently, the uptake

6 L. Shi, M. Quick, Y. Zhao, H. Weinstein, and J. A. Javitch, manuscript inpreparation.

An Intracellular Interaction Network in NSS Proteins

JUNE 20, 2008 • VOLUME 283 • NUMBER 25 JOURNAL OF BIOLOGICAL CHEMISTRY 17693

properties of R60A andD436A (as well as in all followingmuta-tions) were characterized in both backgrounds, whereas ELISAwas done in the HA-WT background and the Zn2� assays weredone in the non-tagged WT DAT background. In both back-grounds, R60A and D436A were functional but displayedmarked decreases in Vmax for [3H]DA uptake, with the largerdecrease observed for R60A (values for HA-tagged constructsare listed in Table 1 with the values for the corresponding non-taggedWTDAT constructs listed in the legend). This decreasein Vmax was accompanied by a substantial decrease (10–20-

fold) in the Km values for [3H]DA uptake (Table 1). ELISAexperiments on R60A and D436A in the HA-DAT backgroundshowed that surface expression of the two mutants was sim-ilar to that observed for HA-DAT itself (Fig. 3a). Thus, thedecrease in Vmax for [3H]DA uptake in R60A and D436Aresulted most likely from impaired function rather thanimpaired surface targeting (Vmax normalized to surfaceexpression shown in Fig. 3b).The effect of Zn2� on [3H]DA uptake in R60A and D436A in

the Zn2�-sensitive WT DAT background agreed with our pre-

FIGURE 1. Ionic interactions among conserved residues in NSS transporters interconnect the N terminus/TM1 region with IL3/TM6 and TM8. a, align-ment of the sequences of N-terminal/TM1, IL3, and TM8 in LeuT, hDAT, hNET, hSERT, hGAT1 hGlyT1b (glycine transporter 1b), and the bacterial transporter Tyt1.Residues corresponding to Arg-51.26, Ser-2676.67, Tyr-2686.68, and Asp-3698.74 in LeuT are highlighted in red. The conserved acidic residues surroundingpositions 1.26 and 8.74 are highlighted in blue, whereas the most conserved residue (1.50 and 8.50) within a TM is in bold. b, two-dimensional schematicrepresentation of DAT. Residues in the intracellular interaction network are highlighted. c, view parallel to the membrane plane of the three-dimensionalstructure of LeuT in which the broken line identified the interconnected region shown in greater detail in panel c. d and e, detail from the encircled region in panelb, turned �20° from the perspective shown in that figure, for LeuT (d) and a homology model of DAT (e). The ionic interactions are identified. Note that thealigned serines at position 6.67 (Ser-267 and Ser-334) in the compared structures are shown with their backbone because the interactions involve the –C � Omoiety.

An Intracellular Interaction Network in NSS Proteins

17694 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 283 • NUMBER 25 • JUNE 20, 2008

diction in that [3H]DA uptake was markedly enhanced, albeitwith a relative larger effect onR60A than onD436A (Fig. 3c). Asfor Y335A (12), the effect was maximal with �10 �M Zn2�. Athigher concentrations, the effect gradually decreased resultingin a bell-shaped dose-response curve. This decrease at highZn2� concentrations correspondsmost likely to the low affinityphase of theWTDAT inhibition curve and, thus, can be attrib-uted to nonspecific effects of Zn2� (12, 13, 15). Altogether,similar phenotypes of the three mutations support the partici-pation of Arg-60 together with Tyr-335 and Asp-436 in theinteraction network described above.

Charge Reversal of Asp-4368.74Partially Rescued the TransportCapacity of R60D—To investigatethe interaction between Arg-60 andAsp-436 we tested whether revers-ing the charges at the two loci(R60D � D436R) could lead to atleast partial functional rescue. Fullrescue would not be expectedbecause the positions of these resi-dues in the interaction networkinvolve them in yet other interac-tions, so that the repositioned argi-nine would not benefit from the sta-bilizing cation-� interaction withTyr-335 (Fig. 1); however, partialrescue would demonstrate theinterdependence of the two resi-dues, consistent with interactionbetween them. The single mutants,R60D and D436R, like the alaninesubstitutions of these residues,markedly lowered Km and Vmax val-ues for [3H]DA uptake (Fig. 4a andTable 1). The function of R60D wasmore impaired than that of D436R,i.e. when corrected for surfaceexpression,Vmax for [3H]DA uptakewas 1.6 and 9.1%, respectively, ofthat observed in HA-DAT (Fig. 4aand Table 1). Note that this differ-ence hints at possible alternativeinteractions with other residues inthe microdomain (see below). TheVmax value for R60D � D435R wasalso markedly reduced; however,the value (�2.9% of HA-DAT) wassignificantly greater than thatof R60D but lower than that forD436R (Fig. 4a). This was notobserved in the “non-rescuing”mutant (R60D � D436A) in whichVmax was even lower than in R60D(Fig. 4a and Table 1).In both R60D and D436R, we

observed enhancement of uptake byZn2� with the most dramatic effect

in R60D (Fig. 4b). In R60D � D436R we also observed anenhancing effect of Zn2�; however, the effect was smaller thanthat in R60D, whereas R60D�D436A showed 6-fold enhance-ment. According to our working hypothesis, the strength ofZn2� enhancement represents a read-out of the transporterfraction initially in the inward facing conformation that can beswitched by Zn2� to the outward facing conformation (13).Therefore, the results suggest that mutation in R60D of Asp-436 to arginine (R60D � D436R) but not to alanine (R60D �D436A) lowers the fraction of transporter in the inward facingconformation and distorts less the conformational equilibrium.

FIGURE 2. Comparative MD study in LeuT of WT compared with Y268A. a, time evolution of the distancesbetween Arg-51.26 and Asp-3698.74 or Ser-2676.67 in a MD trajectory of WT (top) and Y268A mutant (bottom).Note the constant H-bonding distance maintained throughout the 20-ns trajectory for both interactions in theWT, and the fluctuations in and out of H-bonding distance demonstrating the flexibility of this region in themutant (Y268A). At any time after the first 2.5 ns, only one of the distances corresponds to a hydrogen bond,and the alternation in H-bonds distances between Arg-5 and the two ionic neighbors in several time intervals(�9 –13 ns). b, the degree of mobility of the residues in the ionic patch is illustrated by the superposition of sidechain positions explored during the entire 20-ns MD simulation trajectories for the WT compared with theY268A mutant. Note the wider range of conformations experienced by the side chains in the mutant comparedwith the compact bundles in the WT.

An Intracellular Interaction Network in NSS Proteins

JUNE 20, 2008 • VOLUME 283 • NUMBER 25 JOURNAL OF BIOLOGICAL CHEMISTRY 17695

Assessing the Conformation of the Mutant Transporters in aCysteine Reactivity Assay—To obtain a structural read-out forthe conformational state of the mutant transporters weemployed a conformational assay first developed for NET/SERT (27) and subsequently applied by us to DAT (13). Theassay is based on the reactivity of the membrane impermeant,cysteine-reactive, positively charged methanethiosulfonatecompound, MTSET, toward a cysteine introduced in position159 (position 3.53 according to generic nomenclature), which ispredicted to be accessible in the outward facing conformationbut inaccessible in the inward facing conformation of the trans-porter (13, 27). This inference is supported by the fact that thealigned position in LeuT is mostly buried in the crystallizedconformation, which is characterized by a closed extracellulargate (8). In further agreement, the reactivity of Cys-159 withMTSET applied from the externalmilieu is decreased in Y335Abut increased upon application of Zn2�, consistent withdecreased accessibility in the inward facing conformation andpartial rescue into an outward facing conformation (13).We performed the assay in a background in which the only

two reduced cysteines on the extracellular face of the trans-porter (Cys-90 and Cys-306) were mutated (E2C) to eliminateinterference from the MTSET reaction with residues otherthan Cys-159 (15, 23). Because reaction of MTSET with Cys-159 results in transporter inactivation, [3H]DAuptakewas usedas a functional read-out for MTSET reactivity; hence, in agree-ment with our previous observations (13), treatment of cellsexpressing DAT-E2C-I159C with 0.5 mM MTSET caused�30% inhibition of uptake (Fig. 5a). This inhibition wasreduced in both R60D and D436R, which is taken to indicatethat a greater proportion of transporter molecules is in the

inward facing conformation with the extracellular gate closedand thus with diminished reactivity of Cys-159. Addition of 10�M Zn2� during MTSET pretreatment enhanced inhibition inboth R60D and D436R to levels similar to that seen for WTDAT (Fig. 5b), similar to our data previously obtained withY335A (15). Remarkably, MTSET inhibition of the charge-re-

FIGURE 3. Reduction and Zn2� enhancement of DA uptake upon muta-tion of Arg-601.26 or Asp-4368.74 into alanines in DAT. a, similar levels ofsurface expression of HA-DAT-R60A and HA-DAT-D436A as indicated byELISA. Data are mean � S.E. (n � 7 each done in triplicate) with the WT Vmax/expression value being set at 100% in all experiments. b, Vmax for [3H]DAuptake related to the transporter cell surface expression as quantified byELISA on COS7 cells transiently transfected with the indicated DAT-HA con-struct. Data are mean � S.E. (n � 4 each done in triplicate) with the WT Vmax/expression value being set at 100% in all experiments. c, effect of Zn2� on[3H]DA uptake in WT DAT (f) compared with the effect in mutants R60A ()and D436A (�). The data are shown as relative uptake in % of [3H]DA uptakein the absence of Zn2� (mean � S.E., n � 3). The measurements were done intransiently transfected COS7 cells.

TABLE 13H�DA uptake properties measured in COS7 cells transientlyexpressing HA-tagged WT DAT (HA-WT) or the indicated HA-taggedmutantsThe Vmax values for 3H�DA uptake were normalized for surface expression asdetermined by ELISA experiments and are expressed as mean � S.E. in % ofHA-WT. TheVmax forWT-HAwas 4310� 736 fmol/min/105 cells (n� 8). TheKmvalues were calculated from the observed IC50 value found by non-linear regressionanalysis of 3H�DAuptake assays. The S.E. interval for eachKm value is indicated andwas calculated from the pKI � S.E. Data are from at least three independent exper-iments each performed in triplicate. The Km values (in �M) and Vmax values (infmol/min/105 cells), respectively, for 3H�DAuptake in the corresponding non-HA-taggedWTDATwere as follows:WT, 1.14 (1.04–1.24) and 3547� 177; R60A, 0.05(0.04–0.06), and 52.3 � 8.2; D436A, 0.13 (0.10–0.17) and 355 � 13; R60D, 0.08(0.06–0.10), and 74 � 8.2; D436R, 0.13 (0.12–0.14) and 338 � 13; R60D � D436R,0.07 (0.06–0.09) and 184� 9; R60D�D436A, 0.09 (0.09–0.10) and 50� 3.8; E61A,1.27 (1.18–1.36) and 2788 � 623; E61A � D436R, 0.04 (0.02–0.06) and 68.8 � 8.8;E437A, 0.63 (0.57–0.70) and 1850 � 450; D436A � E437A, 0.07 (0.05–0.12) and31.7 � 3.2; R60D � E61R, 0.11 (0.10–0.12) and 425 � 55; R60A � E61R, 0.08(0.04–0.19) and 293 � 62.

Construct Vmax/surface expression Km for DA% of WT HA �M

HA-WT 100 1.61 (134–1.93)HA-R60A 1.23 � 0.03 0.08 (0.06–0.10)HA-D436A 8.30 � 0.56 0.18 (0.14–0.22)HA-R60D 1.60 � 0.09 0.10 (0.07–0.15)HA-D436R 9.13 � 1.20 0.24 (0.18–0.31)HA-R60D � D436R 2.89 � 0.44 0.11 (0.08–0.15)HA-R60D � D436A 0.67 � 0.21 0.08 (0.04–0.14)HA-E61A 88.7 � 8.52 1.78 (1.49–2.13)HA-E61A � D436R 2.12 � 0.56 0.07 (0.05–0.09)HA-E437A 45.4 � 9.24 0.90 (0.68–1.20)HA-D436A � E437A 1.91 � 0.22 0.13 (0.10–0.16)HA-R60D � E61R 7.10 � 0.37 0.06 (0.04–0.09)HA-R60A � E61R 8.28 � 0.76 0.15 (0.14–0.17)

An Intracellular Interaction Network in NSS Proteins

17696 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 283 • NUMBER 25 • JUNE 20, 2008

versed double mutant R60D � D436R, but not of the non-res-cuing control mutant (R60D � D436A), was significantlygreater than that in the individual mutants and close to thatseen for the WT DAT (Fig. 5a), as would be expected for amutation bringing the properties closer to WT. This suggeststhat when the twomutations, R60D and D436R, are combined,WT conformational properties are partially rescued, mostlikely through the formation of an inverted salt bridge.Additional Residues in the Ionic Patch between the N Termi-

nus and TM8—Two highly conserved acidic residues (Glu-61andGlu-437, positions 1.27 and 8.75, respectively, according togeneric nomenclature) are situated adjacent to Arg-6 and Asp-436, respectively, creating a patch of charged residues with onepositively charged residue surrounded by three negativelycharged ones (Figs. 1, a and b, and 6, a and b). The propinquitysuggests that these two residues may also play a role in the

pattern of ionic interactions between the N terminus and TM8.However, mutation of Glu-61 to alanine had almost no impacton transporter function and the effect of Zn2� was similar tothat in WT DAT (Fig. 6, c and d, and Table 1). Mutation ofGlu-437 to alanine decreased the Vmax values for [3H]DAuptake in agreement with previous data (13) (Fig. 6c, Table 1);however, E437A was still inhibited by Zn2� although to aslightly less extent as compared with WT DAT (Fig. 6d). Thus,neither Glu-61 nor Glu-437 appeared vital to transporter func-tion under “normal” circumstances (Fig. 6a).Based on ourDATmodel we reasoned, however, that the two

residues could affect the phenotypes observed upon mutationof Arg-60 and Asp-436. Specifically, althoughmutation of Asp-436 to alanine removed the main component of the TM1-TM8interaction, even a weak interaction between Arg-60 and Glu-

FIGURE 4. Partial rescue of R60D transport capacity by introducingD436R. a, the Vmax value for [3H]DA uptake (normalized to surface expres-sion) is increased in R60D � D436R and decreased in R60D � D436A com-pared with R60D. The Vmax for [3H]DA uptake was normalized to surfaceexpression as quantified by ELISA on COS7 cells transiently expressing theindicated DAT-HA constructs. Data are mean � S.E. (n � 4 each performed intriplicate) with WT Vmax/expression value being set at 100% in all experi-ments. b, effect of 10 �M Zn2� on [3H]DA uptake for the indicated mutationsin WT DAT background transiently expressed in COS7 cells. Data are mean �S.E. of n � 3 each performed in triplicate. * and **, significantly different valuesin a paired t test with p � 0.01 and 0.05, respectively, compared with R60D.

FIGURE 5. Introduction of D436R in R60D increases MTSET accessibility toCys-1593.53 in TM3. a, inhibition of [3H]DA uptake by MTSET (0.5 mM) in COS7cells transiently expressing DAT-E2C-I159C or mutations DAT-E2C-I159C/R60D, DAT-E2C-I159C � D436R, DAT-E2C-I159C � R60D � D436R, or DAT-E2C-I159C � R60D � D436A. Values are mean � S.E. (n � 4 each performed intriplicate). In R60D � D436R the inhibition is significantly different (**, p �0.01, one-way analysis of variance, Newman-Keuls Comparison post hoc test)from that seen for R60D, D436R, and R60D � D436A. b, same as a with MTSET-induced uptake inhibition measured on the same cells in the presence of 10�M Zn2�.

An Intracellular Interaction Network in NSS Proteins

JUNE 20, 2008 • VOLUME 283 • NUMBER 25 JOURNAL OF BIOLOGICAL CHEMISTRY 17697

437 could substitute to a certain extent for the lost interactionbetween Arg-60 and Asp-436 (Fig. 6, a and b). Mutation ofGlu-437 to alanine in the context of D436A supported thisnotion by showing amore profound effect onVmax, hence,Vmaxnormalized for surface expression diminished 4.3-fold for

E437A inHA-DATD436A,whereasVmax for E437A in HA-DAT back-ground diminished 2.2-fold (Fig. 6, cand f, and Table 1). The enhancingeffect of Zn2� was also increased inD436A � E437A as compared withD436A alone (Fig. 6g), supportingthat a higher fraction of transport-ers have assumed the inward facingconformation in this double mutantbecause all possible bridges con-necting theN terminuswithTM8 inthis regionwere eliminated (Fig. 6e).

We next tested Glu-61 for itseffect on D436R. The model sug-gests that mutation of Asp-436 toarginine would generate electro-static repulsions with Arg-60 butthat this might be mitigated by acompensatory interaction betweenGlu-61 and the arginine insertedin position 436 of D436R (Fig. 6e).Therefore, we mutated Glu-61 toalanine in D436R resulting inE61A�D436Rwhere the possibil-ity for ionic interactions betweenthe N terminus and the cytoplas-mic end of TM8 was eliminated(Fig. 6e). As predicted, this muta-tion in D436R reduced Vmax con-siderably in contrast to having noeffect in WT DAT (Fig. 6, c and f).Moreover, the Zn2� effect wasenhanced in E61A � D436R com-pared with D436R (Fig. 6g).If the four charged residues con-

sidered here operate as an ionicpatch with three negative charges(Glu-61, Asp-436, andGlu-437) andone positive charge (Arg-60), partialfunctional rescue of R60D might beobtained also by introducing thepositive charge in position 61 torestore the ionic interactionbetween the N terminus and TM8(Fig. 7a and supplemental Fig. S1)).As predicted, R60D � E61R as wellas R60A � E61R (which was gener-ated to exclude that a rescuing effectwas due to an ionic interactionbetween positions 60 and 61) dis-played higher Vmax than R60D orR60A, and uptake was only weakly

enhanced by Zn2� (Fig. 7, b and c, Table 1), consistent withfunctional rescue.Normal Mode Analysis—The interaction network is likely to

affect the dynamic properties of the whole transporter mole-cule. Specifically,movements of theN terminus and IL3 relative

FIGURE 6. Glu-611.27 and Glu-4378.75 are part of an ionic network surrounding the Arg-601.26–Asp-4368.74

salt bridge. a, DAT model showing acidic residues (Glu-611.27, Asp-4368.74, and Glu-4378.75) surrounding Arg-601.26 in the N terminus at the intracellular end of TM1 (purple) and TM8 (blue). The predicted salt bridgebetween Arg-601.26 and Asp-4368.74 is illustrated by a dashed green line. The other TMs were removed for clarity.b and e, schematics representing the putative interactions between TM1/proximal N terminus (dark gray) andTM8 (light gray) in WT (b) and mutants (e). Acidic residues are represented by a blue letter in a circle, the largelabile arginine by a red letter in an ellipse, and the neutral alanine by a black letter in a circle. The hypotheticalionic interactions are represented by a plain and dashed green line, respectively, whereas the putative electro-static clashes are illustrated by an orange symbol. IP is defined as “interaction potency” between TM1 and TM8rating the predicted strength of ionic interactions in this area of DAT. ��� is the strongest interaction, � and0 correspond to a weak interaction or no interaction, respectively. c and f, Vmax values for [3H]DA uptakenormalized to cell surface expression in the indicated HA-tagged DAT constructs. Data are % of uptake inHA-WT (mean � S.E., n � 3–5). d and g, effect of 10 �M Zn2� on [3H]DA uptake for the indicated mutations in WTDAT background transiently expressed in COS7 cells. Data are % of uptake in the absence of Zn2� (means �S.E., n � 3). * and **, significantly different (p � 0.05 and 0.01, respectively, paired t test) compared with WT DAT(c and d) or to D436A or D436R (f and g).

An Intracellular Interaction Network in NSS Proteins

17698 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 283 • NUMBER 25 • JUNE 20, 2008

to the middle region of TM8 that is close to the substrate bind-ing site might be important, and disruption of the networkmayimpair the capability of the transporter to make reversiblemotions in the gating process. We explored, therefore, the col-lectivemotions in the various constructs of the transporterwithan analysis of the normal modes (28). LeuT was chosen as themodel because application of this analysis to a high resolution

structure is likely to be more reliable, whereas the high degreeof conservation discussed above suggests that the mechanisticdetails would be shared by DAT.We first inspected the low frequency normal modes of WT

LeuT and quantified the movements by measuring the changesin C�-C� distances between Arg-5 and Gln-361, and Tyr-268and Gln-361 (see supplemental Fig. S2 for further details). Toidentify the role of specific interactions in thesemovements, wecompared the results for normalmodes calculated forWTwithY268A and R5A, and used R193A, a residue facing lipid, ascontrol (supplemental Fig. 2B). The dynamic details are visual-ized in the movies provided as supplementary information(supplementary Movies I and II). It is clearly seen from modes11 and 20 in thesemovies that R5A andY268A aremoremobilethan WT on the intracellular side, whereas WT is significantlymore mobile in general (supplementary Movies I and II andsupplemental Fig. S2B). By disrupting the interaction betweenthe N terminus and TM8, the R5Amutation renders the extra-cellular region of the protein less capable of undergoing thedynamic rearrangements coupled to the conformationalchanges at the intracellular end. Mode 21 of WT exhibits thelargest changes of themonitored distances, whereas in R5A andY268A the motions are markedly different (supplemental Fig.S2B).

DISCUSSION

The analysis underlying this study was made possible by thehigh-resolution structural information obtained recently for abacterial member of the NSS family (LeuT) (8–10), which forthe first time enabled the construction of reliable molecularmodels of the mammalian counterparts (11). We focused on adistinctive and conserved residue patch found in the cytoplas-mic half of the protein in both the LeuT structure and in thecognate DAT model (Fig. 1). Given the concordance betweenour previous experimental findings upon mutation of Tyr-335inDAT (12, 13) and the putative role of the interaction networkidentified in this patch in the gating of the transporter, weprobed computationally the role of this conserved tyrosine inthe network. In parallel, we set out to validate experimentallythe existence and function of such a network in DAT.Our comparative MD simulations of WT LeuT and the

mutant Y268A indicated that the cation-� interaction (Tyr-268–Arg-5) and the salt bridge (Arg-5–Asp-369) could mutu-ally stabilize each other (Figs. 1C and 2). Based on the hypoth-esis that Y268A, like the corresponding mutant (Y335A) inDAT, is enriched in the inward-facing conformation (12, 13)and the fact that the Tyr-268–Arg-5 interaction no longerexists in Y268A, we expected the interaction between the Nterminus and IL3 to be significantly weakened or disrupted inthis mutant. This disruption is likely to propagate to the cyto-plasmic end of TM8 through the Arg-5–Asp-369 interaction.This expectation is born out by dynamic analysis, as illustratedby the changes in the corresponding normal modes (supple-mental Fig. S2 and supplementary Movies I and II). Thus, wepropose a functional role for this network connecting the Nterminus, IL3, and TM8, in which alternating conformationsaccount for the opening and closing of the intracellular gate forsubstrate translocation, in a motion that is propagated through

FIGURE 7. Introduction of E61R in R60D or R60A partially rescues thetransport capacity. a, schematics representing the putative network ofinteractions between the TM1/proximal N terminus (dark gray) and TM8 (lightgray) in the indicated DAT mutants (as in Fig. 6). The hypothetical ionic inter-actions are represented by a dashed green line. b, Vmax values for [3H]DAuptake normalized to cell surface expression in the indicated HA-tagged DATconstructs. c, effect of 10 �M Zn2� on [3H]DA uptake for the indicated muta-tions in WT DAT background transiently expressed in COS7 cells. Data aremean � S.E. (n � 3 each performed in triplicate). ***, significantly differentvalues in a t test with p � 0.001 compared with R60D or R60A.

An Intracellular Interaction Network in NSS Proteins

JUNE 20, 2008 • VOLUME 283 • NUMBER 25 JOURNAL OF BIOLOGICAL CHEMISTRY 17699

the entire structure to support the corresponding opening/closing changes in the extracellular end of the transporter.The salt bridge interaction between Arg-60 and Asp-436 in

DAT was probed experimentally by interchanging the two res-idues. This resulted in partial rescue demonstrating the inter-dependence of the two charged residues consistent with aninteraction between them. Note that full rescue would not beexpected from our structural models, because the importantcation-� interaction would not be restored in the revertantmutants. Charge reversal mutations have been used to validatethe existence of salt bridges also in other membrane proteins,and revealed spatial adjacencies even in the absence of three-dimensional structures (29–32). In some cases essentially fullfunctional rescue was observed, as seen, for example, for a saltbridge in the �1b receptor (31), whereas in other studies theobservations were more parallel to our findings with only par-tial functional rescue (29, 30) or more complex phenotypes, asin the GABAB receptor (32).Altogether, our data strongly support the presence of the

interaction network (Fig. 1e) and its impact on the dynamicsrelated to function.Of further interest,mutation of Ser-267wasrecently found to cause inverted Zn2� sensitivity,7 andimpaired transport function was seen upon mutation of Glu428(13), which would disrupt the interaction betweenGlu-428 andTyr-335 (Fig. 1e). Our structural models revealed furthermore,that the microenvironment of the mutated residues is enrichedin putative alternative interaction partners, such as Glu-61 andGlu-437. These appeared to confer some redundancy to theinteraction between TM1 and TM8 by partially compensatingfor the absence of the stabilizing salt bridge in the Asp-436mutants but not in the Arg-60 mutants (Fig. 6, a and e). Addi-tional evidence for the importance of thismicroenvironment ofreinforced interactions was provided by the partial “rescuing”of uptake not only upon interchange ofArg-60 andAsp-436 butalso upon interchange of Arg-60 and Glu-61.This “rescue” (R60D � E61R) was more pronounced than

that in R60D � D436R, most likely because it preserved anarginine in the key position that allows for stabilization of thereversed charge interaction. This idea is supported by theoccurrence of just such a reversed charge combination in one ofthe bacterial transporters from the NSS family, the tyrosinetransporter Tyt1 (33), which possesses an arginine at genericposition 1.27 and a glutamate at 1.26 (Fig. 1a). In ourMD studyof a LeuT-based Tyt1 model, Arg-6 (aligned to Glu61 in DAT)formed not only a salt bridge with Glu-330 (generic position8.70), which is located one turn above the Asp-436 of DAT, butalso a cation-� interaction with Tyr-241 (corresponding toTyr-335 in DAT and Tyr-268 in LeuT) that was in a differentorientation than in LeuT (supplemental Fig. S1) (33). Similarly,we interpret the better rescue seen inDATR60D� E61R as theresult of re-establishing both a salt bridge and a cation-� inter-action (between the N terminus and IL3).Of additional interest, mutation of the residue correspond-

ing to Arg-60 in the homologousGABA transporter-1 (GAT-1)(Arg-44) was shown to impair uptake likely by inhibiting reori-

entation of the unloaded transporter (34), consistent with a rolein closure of the internal gate as suggested here. Furthermore,when the residue following Arg-44 in GAT1, i.e. Asp-45, wasdeleted, the one-residue shift and the presence of an aspartatein position 43 generated a mutation similar to DAT R60D �E61R. This deletion mutant (del45) was functional (34) andthus exhibited a rescue phenotype similar to what we find forR60D � E61R. Taken together, the similarities between theobservations obtained in GAT-1 and the present data obtainedinDAT indicate that the role of Arg-601.26 proposed heremightbe extended to the entire transporter class.By using a comparative normal mode analysis in the context

of the LeuT structure we evaluated the relationship betweenthe dynamic properties of the transporter molecule and theionic interaction between the N terminus and TM8 in the col-lective motions of the transporter. The advantage of thisapproach is that, like conventional time-dependent MD, it isable to identify global structural changes that are critical tosubstrate translocation, but because it is time independent it isnot subject to the same limitations as MD, e.g. high computa-tional cost of long simulations for a large system. Many exam-ples have shown that the collective motions that are function-ally important often follow trajectories along one or a few lowfrequency normal modes (35–37), and that the intrinsic struc-tural flexibility of a proteinmanifested in the normalmodes hasevolved to facilitate functionally important conformationaltransitions (36).Interpreted on this basis, our results fromnormalmode anal-

yses point to a pattern of motions that likely represents an allo-steric response to the binding of the substrate (supplementalFig. S2). The comparative analysis of normal modes 11 and 20was particularly revealing as the motions in mutants and WTare directly comparable (supplementary Movies I and II). Asevidenced by the differences between the correspondingmodesin WT, R5A, and Y268A, the structural context of the relevantlow frequency modes emphasizes the importance that Arg-5,Tyr-268, and the associated network, have for the intrinsic flex-ibility of the transporter. Importantly, the propagation ofmotions represented by these modes connects conformationalchanges in the intracellular portion to the extracellular portionof the protein. The increasedmobility of themutants comparedwithWT in the intracellular side is associated with a reductionin the ability of the extracellular region to complete the corre-spondingmotions seen in theWT (supplementaryMovies I andII). This observation is fully consistent with assigning a func-tional role to the ionic interaction in the WT related to thetransition of the transporter between the outward and inwardfacing conformations.Summarized, the present study provides support for a trans-

port mechanism for NSS proteins that appears unique amongsecondary active ion-coupled transporters. Our mechanisticpredictions clearly differ from those involving movements oftwo symmetrical hairpins reaching from the extracellular andintracellular environments, respectively, that were offered forsodium-coupled glutamate transporters (38). Similarly, ourproposed mechanism differs from the “rocker switch” typemechanism proposed for Lac Permease and the glycerol 3-phos-phate transporter (39, 40). Although the key role of ionic inter-7 Y. Dehnes and J. A. Javitch, unpublished observation.

An Intracellular Interaction Network in NSS Proteins

17700 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 283 • NUMBER 25 • JUNE 20, 2008

actions between the cytoplasmic ends of TM1 and TM8 inter-actions is clearly demonstrated here, the exactmolecular eventsgoverning the continuous disruption and reformation of theseinteractions remains to be determined. It is interesting to spec-ulate that the processmight involve sequential protonation anddeprotonation, e.g. of Asp-436, but demonstration of this andother mechanistic elements requires future efforts.

Acknowledgment—We thank Dorthe Vang Larsen for technicalassistance.

REFERENCES1. Chen, N. H., Reith, M. E., and Quick, M. W. (2004) Pflugers Arch. 447,

519–5312. Gether, U., Andersen, P. H., Larsson, O. M., and Schousboe, A. (2006)

Trends Pharmacol. Sci. 27, 375–3833. Torres, G. E., and Amara, S. G. (2007)Curr. Opin. Neurobiol. 17, 304–3124. Henry, L. K., Defelice, L. J., and Blakely, R. D. (2006) Neuron 49, 791–7965. Zomot, E., Bendahan, A., Quick, M., Zhao, Y., Javitch, J. A., and Kanner,

B. I. (2007) Nature 449, 726–7306. Jardetzky, O. (1966) Nature 211, 969–9707. Rudnick, G. (2006) Handb. Exp. Pharmacol. 175, 59–738. Yamashita, A., Singh, S. K., Kawate, T., Jin, Y., and Gouaux, E. (2005)

Nature 437, 215–2239. Zhou, Z., Zhen, J., Karpowich, N. K., Goetz, R. M., Law, C. J., Reith, M. E.,

and Wang, D. N. (2007) Science 317, 1390–139310. Singh, S. K., Yamashita, A., and Gouaux, E. (2007) Nature 448, 952–95611. Beuming, T., Shi, L., Javitch, J. A., andWeinstein, H. (2006)Mol. Pharma-

col. 70, 1630–164212. Loland, C. J., Norregaard, L., Litman, T., and Gether, U. (2002) Proc. Natl.

Acad. Sci. U. S. A. 99, 1683–168813. Loland, C. J., Granas, C., Javitch, J. A., and Gether, U. (2004) J. Biol. Chem.

279, 3228–323814. Goldberg, N. R., Beuming, T., Soyer, O. S., Goldstein, R. A.,Weinstein, H.,

and Javitch, J. A. (2003) Eur. J. Pharmacol. 479, 3–1215. Loland, C. J., Norregaard, L., and Gether, U. (1999) J. Biol. Chem. 274,

36928–3693416. Tirion, M. M. (1996) Phys. Rev. Lett. 77, 1905–190817. Humphrey, W., Dalke, A., and Schulten, K. (1996) J. Mol. Graph. 14,

33–38

18. Phillips, J. C., Braun, R., Wang, W., Gumbart, J., Tajkhorshid, E., Villa, E.,Chipot, C., Skeel, R. D., Kale, L., and Schulten, K. (2005) J. Comput. Chem.26, 1781–1802

19. Sotomayor, M., and Schulten, K. (2004) Biophys. J. 87, 3050–306520. Sali, A., and Blundell, T. L. (1993) J. Mol. Biol. 234, 779–81521. Dunbrack, R. L., Jr., and Cohen, F. E. (1997) Protein Sci. 6, 1661–168122. Suhre, K., and Sanejouand, Y. H. (2004) Nucleic Acids Res. 32,

W610–W61423. Gallivan, J. P., and Dougherty, D. A. (1999) Proc. Natl. Acad. Sci. U. S. A.

96, 9459–946424. Slutsky, M. M., and Marsh, E. N. (2004) Protein Sci. 13, 2244–225125. Norregaard, L., Frederiksen, D., Nielsen, E. O., and Gether, U. (1998)

EMBO J. 17, 4266–427326. Sorkina, T.,Miranda,M., Dionne, K. R., Hoover, B. R., Zahniser, N. R., and

Sorkin, A. (2006) J. Neurosci. 26, 8195–820527. Chen, J. G., and Rudnick, G. (2000) Proc. Natl. Acad. Sci. U. S. A. 97,

1044–104928. Cui, C., and Bhar, I. (eds) (2006) Normal Mode Analysis: Theory and

Applications to Biological and Chemical Systems, Chapman & Hall/CRC,Boca Raton, FL

29. Zhou, W., Flanagan, C., Ballesteros, J. A., Konvicka, K., Davidson, J. S.,Weinstein, H., Millar, R. P., and Sealfon, S. C. (1994)Mol. Pharmacol. 45,165–170

30. Sealfon, S. C., Chi, L., Ebersole, B. J., Rodic, V., Zhang, D., Ballesteros, J. A.,and Weinstein, H. (1995) J. Biol. Chem. 270, 16683–16688

31. Porter, J. E., and Perez, D. M. (1999) J. Biol. Chem. 274, 34535–3453832. Binet, V., Duthey, B., Lecaillon, J., Vol, C., Quoyer, J., Labesse, G., Pin, J. P.,

and Prezeau, L. (2007) J. Biol. Chem. 282, 12154–1216333. Quick,M., Yano, H., Goldberg, N. R., Duan, L., Beuming, T., Shi, L.,Wein-

stein, H., and Javitch, J. A. (2006) J. Biol. Chem. 281, 26444–2645434. Bennett, E. R., Su, H., and Kanner, B. I. (2000) J. Biol. Chem. 275,

34106–3411335. Hayward, S., Kitao, A., and Go, N. (1995) Proteins 23, 177–18636. Ma, J. (2005) Structure 13, 373–38037. Tama, F. (2003) Protein Pept. Lett. 10, 119–13238. Boudker, O., Ryan, R. M., Yernool, D., Shimamoto, K., and Gouaux, E.

(2007) Nature 445, 387–39339. Abramson, J., Smirnova, I., Kasho, V., Verner, G., Kaback, H. R., and Iwata,

S. (2003) Science 301, 610–61540. Huang, Y., Lemieux, M. J., Song, J., Auer, M., and Wang, D. N. (2003)

Science 301, 616–620

An Intracellular Interaction Network in NSS Proteins

JUNE 20, 2008 • VOLUME 283 • NUMBER 25 JOURNAL OF BIOLOGICAL CHEMISTRY 17701