the binding pocket of vSGLT
Membrane co-transport proteins that use a five-helix inverted
repeat motif have recently emerged as one of the largest structural
classes of secondary active transporters1,2. However, despite many
structural advances there is no clear evidence of how ion and sub-
characterization and a new crystal structure of the inward-open con-
formation at a resolution of 2.7A˚. Our data show that sodium exit
causesa reorientation of transmembrane helix 1 that opens an inner
gate required for substrate exit, and also triggers minor rigid-body
movements in two sets of transmembrane helical bundles. This
cascade of events, initiated by sodium release, ensures proper timing
of ion and substrate release. Once set in motion, these molecular
tose readily to enter the intracellularspace. Additionally, we identify
is essential in the coupling of co-transport.
Secondary active transporters harness the energy stored in electro-
chemical gradients to drive the accumulationof specific solutes across
cell membranes. This task is accomplished by the alternating-access
mechanism, in which the substrate-binding site is first exposed to one
side of the membrane and, on ion and substrate binding, a conforma-
tional change exposes the transported solute to the opposite face,
where it is released3. Sodium/glucose co-transporters are prototypes
of secondary active transporters that drive the accumulation of sugars
human physiology, where mutations in their genes are responsible for
severe congenital diseases4and are the molecular targets for drugs to
treat diabetes and obesity5.
There has been a recent surge of work on crystal structures6–11dis-
playing the five-helix inverted repeat motif. These are referred to as the
a wide range of substrates and differ in the number and type of driving
through comparisons of these diverse structures1,2,13. Despite sharing a
common set of ten core transmembrane segments, the lack of sequence
similarity and the chemical diversity of the transported substrates pre-
vents the complete understanding of the mechanistic basis of transport.
This hurdle is being surmounted as multiple structures of the same
ition from outward- to inward-facing conformations17. However, an
atomic-level understanding of sodium-coupled substrate co-transport,
necessary to explain the dynamics of alternating access, is still absent.
out molecular dynamics simulations on the galactose-bound inward-
occluded conformation of vSGLT6embedded in a lipid bilayer18. All
sodium co-transporters of the LeuT superfamily share a common
sodium-binding site termed the Na2 site. During the transition to
theinward-facing conformation, transmembrane helix (TM) 8,which
a less favourable Na2 site that facilitates Na1release1,17(Fig. 1b). Na1
modelled at this site is loosely coordinated by the carbonyl oxygens of
Ile65 (3.3A˚), Ala361 (3.2A˚) and the side-chain hydroxyl of Ser365
(3.1A˚). The carbonyl oxygen of Ala62 (3.6A˚) and the side-chain
hydroxyl of Ser364 (3.6A˚) are also in close proximity (Fig. 1b).
Previous molecular dynamics simulations performed on vSGLT19
and Mhp117indicatedthatNa1quicklyleavestheNa2 site.Oursimu-
lations indicate that Na1exits the Na2 site after 9ns (Fig. 2a) and
interacts with the hydrophilic pore-lining residue Asp189 on TM5
during exit. The importance of Asp189 was highlighted in a previous
simulation19and in biochemical studies on hSGLT120. All three
molecular dynamics simulations indicate that Na1exits the trans-
porter before substrate exit; however, additional conformational
changes are required to release the occluded galactose.
Intheinward-occluded structure, galactoseislocated halfway across
the membrane (Figs 1a and 2b), where it is coordinated by extensive
oftheseresiduesform twohydrophobicgatesblockinggalactose exitto
the intracellular and extracellular spaces. Our molecular dynamics
simulations show that as Na1exits the Na2 site, galactose undergoes
significant fluctuations within the binding pocket. At 52ns, Tyr263
adopts a new and stable rotamer conformation that expands the exit
pathway (between 52 and 110ns), permitting the sugar to leave the
binding site (Figs 2 and 3). After sugar release (,110ns), Tyr263
returns to the original conformation.
To test the hypothesis that Na1release stimulates an alternative con-
lation in which the sodium was lightly restrained in the Na2 site. Under
suggeststhatsodium release drivesconformational changesthatdisrupt
the Na2 site and Tyr263 are central to the transport mechanism.
makes possible the accurate determination of the binding free energy
profile through the use of umbrella sampling along the exit pathway
coupled with weighted histogram analysis21(Fig. 3, inset). After Na1
release, galactose is weakly bound to vSGLT with a minimal energy
barrier of ,2kcalmol21, resulting from the interaction of the sugar
with residues Asn64, Ser66, Glu68 and Gln69 on TM1. Asn64 is of
particular interest because it is located in the unwound segment of
TM1andhashydrogen bondswiththe innergate residueTyr263and
the O2 hydroxyl of galactose linking the Na2 site with the galactose
site. Thus, the interactions of Asn64 with Tyr263 and galactose may
be critical to the transport mechanism6.
To test the importanceof theseinteractions, weperformedmolecu-
lar dynamics simulations and sodium-dependent transport assays on
1Department of Physiology, University of California, Los Angeles, Los Angeles, California 90095-1759,USA.2Department of Biological Sciences, University of Pittsburgh, Pittsburgh, Pennsylvania 15260,
USA.3Department of Computational and Systems Biology, University of Pittsburgh, Pittsburgh, Pennsylvania 15260, USA.
*These authors contributed equally to this work.
9 8 8 | N A T U R E | V O L 4 6 8 | 1 6 D E C E M B E R 2 0 1 0
Macmillan Publishers Limited. All rights reserved
the simulation. The failure of Na1to unbind prevents conformational
changes in the unwound segment of TM1, and Tyr263 remains in the
blocked orientation (Supplementary Fig. 2a). In agreement with the
simulation, sodium-dependent transport assays on the Asn64Ala
mutant show no activity (Fig. 2c).
To explore the role of Asn64 further, we tested Asn64Gln and
Asn64Ser, which, in principal, are both capable of maintaining the
prevented simulation as the result of substantial steric clashes, which
correlated well with a lack of transport (Fig. 2c). The model of the
but not with the O2 hydroxyl of galactose (4.3A˚). In the simulation,
Asn64Ser releases Na1from the Na2 site at 25ns, and Tyr263 tran-
siently adopts the alternative rotamer conformation before returning
to its original position, preventing galactose exit (Supplementary Fig.
2c). Similarly, simulation of the Tyr263Phe mutant shows that Na1
unbinds at 10ns, but unlike tyrosine, phenylalanine never adopts a
conformationcompatiblewith galactoseexit (Supplementary Fig. 2b).
Longer simulations may reveal galactose release in these mutants,
however, the transport assays and simulation data demonstrate that
robust transport requires precise orientation of Asn64 to stabilize
galactose and the gating residue Tyr263.
Although the molecular dynamics simulations and biochemical
studies demonstrate a physical link between the Na2 site and the
substrate, global details regarding the inward-open conformation
(devoid of both ligands) remain elusive. To address this issue, we
determined thestructureof vSGLTin the inward-openconformation.
Crystals, in the absence of ligands, for both the wild-type protein and
the inactive Lys294Ala mutant6were obtained. Both crystals had the
same overall configuration, but the mutant crystals diffracted to a
higher resolution (2.7A˚; see Methods).
posed of 14 transmembrane helices, ten of which comprise the core
domain. TM1–TM5 and TM6–TM10 are related by an approximate
two-fold symmetry axis through the centre of the membrane plane.
The inward-occluded and the inward-open structures have a similar
overall fold with a r.m.s.d. of 1.2A˚. However, there are distinct struc-
changes resulting from the release of ligands (Fig. 4). With the excep-
tion of TM1, superimpositions of individual helices reveal that the
occluded-to-open transition occurs by rigid-body movements of sub-
domains (Supplementary Figs 3 and 4). Consistent with the recent
Figure 1 | Structures and overlay of the inward-open and inward-occluded
conformations. a, Thecoredomainoftheinward-openconformation(TM1–
TM10)iscolouredby specifichelixbundlesinvolvedin thetransition from the
inward-occluded to the inward-open conformation. The ‘hash motif’ formed
from TM3, TM4, TM8 and TM9 is blue; the ‘sugar bundle’ formed from TM2,
TM6 and TM7 is green; TM1 is red; and TM5 and TM10 are magenta. The
periphery helices (TM21, TM11, TM12 and TM13) are yellow. Atoms are
displayed in ball-and-stick form with oxygen coloured red and nitrogen
coloured blue. Inset, an overlay of the inward-open (colour) and inward-
occluded (grey) conformations illustrating the coordination at the Na2 and
galactose-binding sites. b, c, Overlay of the inward-open and inward-occluded
conformations with thesame colouringas in a. Conformational changes in the
inward-open structure reveals a ,13u kink in the unwound segment of TM1
that prevents sodium coordination at the Na2 site (b). In the absence of
galactose, the galactose-binding residue Asn64 hydrogen-bonds to Glu88 and
Tyr263, maintaining an open pathway from the intracellular space to the
substrate-binding site (c).
0 20 4060
Percentage of control
Figure 2 | Mechanism of galactose release. a, Sodium and galactose exit
vSGLT. The root mean squared deviation (r.m.s.d.) of Na1(green) rapidly
is shown in the conformation observed in the inward-occluded structure6, in
52ns (shown on the right), Tyr263 adopts a rotamer conformation that
expands the exit pathway. c, D-galactose uptake by wild-type and vSGLT
mutants in proteoliposomes. Results are expressed as percentage uptake in
Asn64Gln and Tyr263Phe severely impair sodium-dependent transport.
Error bars, s.e.m. WT, wild type.
1 6 D E C E M B E R 2 0 1 0 | V O L 4 6 8 | N A T U R E | 9 8 9
Macmillan Publishers Limited. All rights reserved
assignment for Mhp117, the hash motifs, formed from TM3 and TM4
and their inverted repeat equivalents, TM8 and TM9, align with a
bundle for the extensive side-chain interactions with galactose, and
these regions superimpose with a r.m.s.d. of 0.5A˚. This new inward-
open structure of vSGLT is more similar to the recent inward-facing
conformation structure of Mhp1 than is the previous structure of
vSGLT. Details of this structural analysis are in Methods.
ture is presumably triggered by sodium release from the Na2 site and
the alteration in the hydrogen-bonding network surrounding the
unwound segment of TM1. In particular, the intracellular half of
TM1 flexes ,13u, modifying the coordination of Asn64 (Figs 1b
and 4a). IntheabsenceofbothgalactoseandNa1,Asn64coordinates
and O3 hydroxyls of galactose. This new conformation of TM1 is
further stabilized by hydrogen bonds between the Na2-site residue
Ser365 and Glu68 on the unwound segment of TM1 (Supplemen-
tary Fig. 5). When viewed from the intracellular side, each domain
accessibility cavity by ,1,400A˚3(Fig. 4). This 6u relative rotation
probably disrupts protein–substrate coordination and permits water
which an increase in the number of water molecules in the substrate-
binding site is observed after sodium release (Supplementary Fig. 6).
ening galactose in the pocket and ultimately assisting in its release
(Supplementary Fig. 6 and Supplementary Movie 1).
from vSGLT. The transition from the outward- to the inward-
exit, the carbonyl oxygens of the ion-coordinating residues Ile65 and
Ala62 undergo a conformational change in the unwound segment of
that movement of TM1 disrupts the hydrogen bond between Asn64
and Tyr263, allowing the side chain of Tyr263 to adopt a new con-
formation that opens a pathway to the intracellular space (Fig. 2b).
Additional rigid-body movements widen the intracellular cavity,
site to enhance exit and prevent rebinding (Fig. 4).
It is likely that the reaction scheme described here for vSGLT is
broadly used by all sodium-dependent members of the LeuT super-
family, because the Na2 site, the hydrophobic gates and the unwound
a single sodium-binding site, the Na2 site directly interacts with the
substrate through polar residues—Asn64 in vSGLT and Gln42 in
Mhp1—located on the unwound segment of TM1. For proteins that
the additional site (the Na1 site) is positioned on the opposite side of
the unwound helix from the Na2 site. Interactions between the Na1
and Na2 sites are mediated by the unwound segment of TM1, and the
sodium at the Na1 site is directly coordinated to the substrate8,12.
protein, that regulates sodium and substrate release. This primary
structural feature coupling sodium and substrate co-transport has
biology and for developing strategies to manipulate the alternating-
access mechanism therapeutically.
Energy (kcal mol–1)
05 10 1520 25–5
Figure 3 | The potential of mean force for galactose unbinding. Energy of
galactose binding to vSGLT in the absence of Na1. Umbrella sampling along
the natural, equilibrium pathway shown (inset) was used to determine the
X-ray structure is shown along the x axis. The coloured arrows correspond to
at 5A˚, which corresponds to galactose interaction with residues in the kink
region of TM1. Error bars were determined by splitting the production data
into four equal sets, computing the energy profile for each set, and then
the 16 positions marked with points.
Figure 4 | Conformational changes in the transition from the inward-
occluded to the inward-open structure. a, TM1 superimposed between the
inward-open (red) and inward-occluded (grey) structures, showing a ,13u
occluded (grey) conformations. Rigid-body rotations of the hash motif and
sugar bundle by 3u in opposite directions expose the substrate-binding site to
the intracellular environment. c, Accessibility cavity of the inward-occluded
conformation is coloured blue. d, Accessibility cavity of the inward-open
conformation is coloured gold. The conformational changes from TM1, hash
of the inward-open conformation, aiding galactose release.
9 9 0 | N A T U R E | V O L 4 6 8 | 1 6 D E C E M B E R 2 0 1 0
Macmillan Publishers Limited. All rights reserved
Moleculardynamicssimulations. ThevSGLTmonomer(ProteinDataBank ID,
membrane bilayer using the OPM24and CHARMM-GUI18software packages.
Simulations were carried out using NAMD25with the CHARMM27 parameter
set in a 150mM NaCl bath. See Methods for more details.
Protein expression and purification. Plasmids carrying wild-type or mutant
and a Superdex 200 column (size-exclusion chromatography). See Methods for
Transport assays. We generated proteoliposomes by reconstituting purified
vSGLT protein with sonicated lipid at a protein/lipid ratio of 1:200. We measured
transport activity by monitoring the uptake of D-galactose, with14C-D-galactose
(K1replacing Na1). See Methods for more details.
Crystallization and data collection. We concentrated purified wild-type and
Lys294Ala protein to ,13mgml21and grew crystals by the hanging-drop
vapour diffusion method using the Mosquito nanolitre-dispensing robot. Data
and scaled,and phases were calculated by molecularreplacement.The model was
Full Methods and any associated references are available in the online version of
the paper at www.nature.com/nature.
Received 11 May; accepted 12 October 2010.
Published online 5 December 2010.
1.Abramson, J. & Wright, E. M. Structure and function of Na1-symporters with
inverted repeats. Curr. Opin. Struct. Biol. 19, 425–432 (2009).
Krishnamurthy, H.,Piscitelli,C.L. & Gouaux, E. Unlocking the molecular secrets of
sodium-coupled transporters. Nature 459, 347–355 (2009).
Wright, E. M., Hirayama, B. A. & Loo, D. F. Active sugar transport in health and
disease. J. Intern. Med. 261, 32–43 (2007).
Isaji, M. Sodium-glucose cotransporter inhibitors for diabetes. Curr. Opin. Investig.
Drugs 8, 285–292 (2007).
Faham, S. et al. The crystal structure of a sodium galactose transporter reveals
mechanistic insights into Na1/sugar symport. Science 321, 810–814 (2008).
Weyand, S. et al. Structure and molecular mechanism of a nucleobase-cation-
symport-1 family transporter. Science 322, 709–713 (2008).
Ressl, S., Terwisscha van Scheltinga, A. C., Vonrhein, C., Ott, V. & Ziegler, C.
Molecular basis of transport and regulation in the Na1/betaine symporter BetP.
Nature 458, 47–52 (2009).
Fang, Y. et al. Structure of a prokaryotic virtual proton pump at 3.2A˚resolution.
Nature 460, 1040–1043 (2009).
10. Gao, X. et al. Structure and mechanism of an amino acid antiporter. Science 324,
11. Shaffer, P. L., Goehring, A., Shankaranarayanan, A. & Gouaux, E. Structure and
mechanism of a Na1-independent amino acid transporter. Science 325,
12. Yamashita, A., Singh, S. K., Kawate, T., Jin, Y. & Gouaux, E. Crystal structure of a
bacterial homologue of Na1/Cl2-dependent neurotransmitter transporters.
Nature 437, 215–223 (2005).
flux by symmetrical transporters. Physiology (Bethesda) 24, 377–386 (2009).
14. Zhou, Z. et al. LeuT-desipramine structure reveals how antidepressants block
neurotransmitter reuptake. Science 317, 1390–1393 (2007).
15. Gao, X. et al. Mechanism of substrate recognition and transport by an amino acid
antiporter. Nature 463, 828–832 (2010).
LeuT in an open-to-out conformation. Science 322, 1655–1661 (2008).
the sodium-hydantoin transporter Mhp1. Science 328, 470–473 (2010).
18. Jo, S., Kim, T., Iyer, V. G. & Im, W. CHARMM-GUI: a web-based graphical user
interface for CHARMM. J. Comput. Chem. 29, 1859–1865 (2008).
19. Li, J. & Tajkhorshid, E. Ion-releasing state of a secondary membrane transporter.
Biophys. J. 97, L29–L31 (2009).
20. Quick, M., Loo, D. D. & Wright, E. M. Neutralization of a conserved amino acid
residue in the human Na1/glucose transporter (hSGLT1) generates a glucose-
gated H1channel. J. Biol. Chem. 276, 1728–1734 (2001).
21. Kumar, S., Bouzida, D., Swendsen, R. H., Kollman, P. A. & Rosenberg, J. M. The
weighted histogram analysis method for free-energy calculations on
biomolecules. 1. The method. J. Comput. Chem. 13, 1011–1021 (1992).
22. Yernool, D., Boudker, O., Jin, Y. & Gouaux, E. Structure of a glutamate transporter
homologue from Pyrococcus horikoshii. Nature 431, 811–818 (2004).
23. Hunte, C. et al. Structure of a Na1/H1antiporter and insights into mechanism of
action and regulation by pH. Nature 435, 1197–1202 (2005).
in membranes: a computational approach. Protein Sci. 15, 1318–1333 (2006).
25. Kale, L. et al. NAMD2: Greater scalability for parallel molecular dynamics.
J. Comput. Phys. 151, 283–312 (1999).
Supplementary Information is linked to the online version of the paper at
Abramson, Wright and Grabe labs for useful discussions and for reading the
this work, S. Iwata for advance release of the Mhp1 coordinates (Protein Data Bank ID,
2X79), and R. Roskies for assistance with the computations. Simulations were carried
out through a TeraGrid grant at the Pittsburgh Supercomputing Center and the Texas
Advanced Computing Center. This work was supported by NIH grants GM078844
(J.A.), RGY0069 (J.A.) and DK19567 (E.M.W.), and a grant from the Human Frontier
Science Program (J.A.). M.G. is an Alfred P. Sloan Research Fellow.
Author Contributions Experiments were carried out and diffraction data collected by
A.W., V.C. and J.A. Simulations were carried out by S.C. Data were analysed and the
manuscript was prepared by all authors.
Author Information Coordinates and structure factors of the inward-open vSGLT
structure have been deposited in the Protein Data Bank under accession number
2XQ2. Reprints and permissions information is available at www.nature.com/reprints.
The authors declare no competing financial interests. Readers are welcome to
comment on the online version of this article at www.nature.com/nature.
Correspondence and requests for materials should be addressed to J.A.
(firstname.lastname@example.org) or M.G. (email@example.com).
1 6 D E C E M B E R 2 0 1 0 | V O L 4 6 8 | N A T U R E | 9 9 1
Macmillan Publishers Limited. All rights reserved
METHODS Download full-text
Molecular dynamics simulations. Initially, TM21 wasremoved and sixmissing
residues in the TM4–TM5 loop were added with the loop modelling routine in
Modeller26. Residues 53–547 were then embedded in a membrane andsolvated in
a hexagonal box approximately 96396384A˚3in volume for a total of 63,000
atoms. Electroneutrality was enforced with the addition of 150mM NaCl.
Simulations were carried out with the CMAP corrected27CHARMM27 para-
meter set and the TIP3P water model. VMD28and MATLAB were used for
visualization and analysis. The system was minimized using conjugate gradient
minimization and heated to 310K using Langevin dynamics with a 10-ps21
damping coefficient. An initial 300-ps equilibration using the NVT ensemble
atoms were constrained in a harmonic potential with a force constant of
k510.0kcalmol21A˚22. We then switched to an NPT ensemble, and the
restraints on the water molecules and heavy side-chain atoms were gradually
removed in five steps over 1.5ns. All remaining restraints were removed in six
steps over the next 1.8ns. Finally, 10ns of restraint-free simulation was run. All
period and 100-fs decay was used to set the pressure to 1atm. Hydrogen bond
lengthswere constrainedwithSHAKE29, anda 2-fstimestepwasused.A 10A˚van
der Waals cut-off was used along with the particle mesh Ewald method for the
spherical harmonic potential using the distance from the Na1to the centre of mass
produced minimaldistortions in the protein. A 10-nsequilibrationwas runbefore a
entire proteinand3.0A˚forthe restrained-Na1simulation.
Potential of mean force calculation. The potential of mean force (PMF) was
calculated using umbrella sampling with WHAM21. We extracted 69 snapshots
along the pathway and held each configuration in a harmonic potential
(k57.0kcalmol21A˚22) with a resting length equal to the z component of the
distance between the galactose COM and the binding-site COM defined by the
binding-site residues. Two nanosecond trajectories were run for each umbrella,
and the last 1,800ps were used for calculating the PMF. Splitting the trajectories
into two equal parts (200–1,000ps and 1,200–2,000ps) and computing separate
PMFs revealed that the total PMF is well converged.
Protein purification. vSGLT proteins were cloned, expressed and purified as
previously described6,30–32. Briefly, the plasmids were transformed into the
TOP10 cell line expressed to OD6001.8 and induced with 0.66mM L-arabinose
for 4h at 29uC. Cell membranes were isolated, solubilized with 2% decyl-b-D-
maltopyranoside and affinity-purified on a Ni-NTA column. The sample was
further purified by size-exclusion chromatography (Superdex 200) and washed
with crystal buffer (20mM Tris (pH 7.5), 25mM NaCl, 0.174% decyl-b-D-
maltopyranoside) in a 50-kDa Amicon filter unit.
Transport assays. Mutants were created with the QuikChange method and puri-
fied as above. vSGLT protein was reconstituted in 150mM KCl, 10mM Tris/
Hepes (pH 8.0), 1mM DTT, 1mM Na2EDTA, 1mM CaCl2, 1mM MgCl2and
0.5% decyl-b-D-maltopyranoside, with 1.2mgml21sonicated lipid (90mg asolectin
at 4uC. The proteoliposomes were collected and washed twice by centrifugation.
in liquid nitrogen.
Uptake of D-galactose (88mM) with14C-D-galactose tracer into proteolipo-
somes was measured for 18min at 22uC in the presence or absence of 100mM
Na1(K1replacing Na1) as described previously6,30. Proteoliposomes were col-
by scintillation counting. Results are expressed as the mean6s.e.m. of three
determinations and three trials.
Crystallization. Protein was concentrated to ,13mgml21before plating.
solution containing 0.1M MES (pH 6.5), 4% MPD and 9–13% PEG400, and
tridecyl-b-D-maltopyranoside to a final concentration of 0.0017% as an additive.
Before freezing, crystals were cryoprotected using a solution containing 30%
PEG400 and 0.174% decyl-b-D-maltopyranoside.
Data processing, phasing and refinement. Data was collected at 1.0A˚on cryo-
using an anisotropy correction server34(resolution cut-offs: a53.1A˚, b52.7A˚
and c52.8A˚). Phases were calculated by molecular replacement (PHASER35)
using the original vSGLT structure as a search model. The model was built in
COOT36and refined using PHENIX37and BUSTER38using non-crystallographic
symmetry (NCS) and TLS refinement restraints. There are two molecules per
to an Rwork/Rfreevalue of 25.1/27.4. The Ramachandran statistics shown areas
follows: 95.5% of the residues lie in the preferred region, 4.3% lie in the allowed
region and 0.2% are outliers.
The 2Fo–Fcmaps contained three elongated features having a maximal peak
height of 3s. These attributes were interpreted and assigned as PEG molecules.
Two are located at theperiphery, whereas the third is near theNa2siteasobserved
The Lys294Alaproteincrystalsdiffract tohigher resolutionthanthewild-type
crystals. Data from four wild-typecrystals were collected andmerged to achieve a
Lys294Ala mutant model and no significant peaks were observed. The
Lys294Alamodel wasfurther refined usingPHENIXto yieldan Rwork/Rfreevalue
We note that although refinement was carried out with data subject to aniso-
tropic correction, as described above, the deposited data has not been treated.
Figures were created from the A-chain protomer using PYMOL39.
Structural comparison of vSGLT with Mhp1. Superpositions of the inward-
occluded (Protein Data Bank ID, 3DH4) and inward-open conformations of
vSGLT with the inward-facing conformation of Mhp1 (Protein Data Bank ID,
near the substrate- and ion-binding sites. The Na2-site helices (TM1, TM5 and
TM8) of the inward-open conformation of vSGLT have a closer fit to Mhp1
(r.m.s.d., 2.2A˚) than the inward-occluded conformation (r.m.s.d., 2.6A˚); thus,
the inward-open vSGLT structure more closely resembles the structure of Mhp1.
26. Sali, A. & Blundell, T. L. Comparative protein modelling by satisfaction of spatial
restraints. J. Mol. Biol. 234, 779–815 (1993).
27. MacKerell, A. D. Jr, Feig, M. & Brooks, C. L. III. Improved treatment of the protein
backbone in empirical force fields. J. Am. Chem. Soc. 126, 698–699 (2004).
28. Humphrey, W., Dalke, A. & Schulten, K. VMD: visual molecular dynamics. J. Mol.
Graph. 14, 33–38 (1996).
29. Ryckaert, J. P., Ciccotti, G. & Berendsen, H. J. C. Numerical integration of the
Cartesian equations of motion of a system with constraints: molecular dynamics
of n-alkanes. J. Comput. Phys. 23, 327–341 (1977).
30. Turk, E. et al. Molecular characterization of Vibrio parahaemolyticus vSGLT: a
model for sodium-coupled sugar cotransporters. J. Biol. Chem. 275,
secondary structure of functionally active vSGLT, the Vibrio sodium/galactose
cotransporter. Biochemistry 45, 1470–1479 (2006).
32. Veenstra, M., Lanza, S., Hirayama, B. A., Turk, E. & Wright, E. M. Local
conformational changes in the Vibrio Na1/galactose cotransporter. Biochemistry
43, 3620–3627 (2004).
33. Otwinowski, Z. & Minor, W. in Macromolecular Crystallography (eds Abelson, J. N.,
34. Strong,M.etal.Towardthestructuralgenomics ofcomplexes:crystalstructureof
USA 103, 8060–8065 (2006).
35. Qian, B. et al. High-resolution structure prediction and the crystallographic phase
problem. Nature 450, 259–264 (2007).
36. Emsley, P. & Cowtan, K. Coot: model-building tools for molecular graphics. Acta
Crystallogr. D 60, 2126–2132 (2004).
37. Adams, P. D. et al. Recent developments in the PHENIX software for automated
X-ray data can deliver improved models for most PDB entries. Acta Crystallogr. D
65, 176–185 (2009).
39. DeLano, W. L. PyMOL Molecular Viewer Æhttp://www.pymol.orgæ (2002).
Macmillan Publishers Limited. All rights reserved