, 1729 (2007);
et al. Ivo Melcák,
Pore Diameter by Intermolecular Sliding
Structure of Nup58/45 Suggests Flexible Nuclear
www.sciencemag.org (this information is current as of April 24, 2009 ):
The following resources related to this article are available online at
version of this article at:
including high-resolution figures, can be found in the online
Updated information and services,
can be found at:
Supporting Online Material
, 16 of which can be accessed for free:
cites 26 articles
15 article(s) on the ISI Web of Science.
This article has been
6 articles hosted by HighWire Press; see:
This article has been
This article appears in the following
in whole or in part can be found at:
permission to reproduce
of this article or about obtaining
Information about obtaining
registered trademark of AAAS.
is a Science2007 by the American Association for the Advancement of Science; all rights reserved. The title
CopyrightAmerican Association for the Advancement of Science, 1200 New York Avenue NW, Washington, DC 20005.
(print ISSN 0036-8075; online ISSN 1095-9203) is published weekly, except the last week in December, by the Science
on April 24, 2009
tide groups include parts of the adenine ring and
the phosphate-distal face of the ribose moiety.
AMP- and ATP-bound structures, suggesting that
the nucleotide-binding face of the AMPK regu-
latory trimer is unlikely to function as the site of
activation modulated by adenylate binding. The
primary differences between AMP- and ATP-
bound forms of the heterotrimer lie within the
phosphate tunnel. One substantial difference be-
tween the ATP and AMP complexes is in the sur-
face electrostatic potentials at the distal exit of the
phosphate tunnel, the putative kinase-interaction
face (Fig. 2, D to G). This suggests the possibility
from the chargedifference between the mono- and
triphosphate groups of AMP and ATP.
Prior studies have identified a number of mu-
to impaired function of AMPK, primarily in the g
gS247 (N488I in human g2) (33), which is found
within the dimer-of-trimers interface region. How-
ever, the large majority of functionally important
mutations, which include changes to residues
gR290 (human g2 R531) (34), gR141 (human g2
H383) (32), gV56 (human g2 R302) (35), and
gT162 (human g2 T400N) (33), are all found lin-
ing the interior surface of the phosphate tunnel
(Fig. 3, A and B). Residues gR290 and gR141 co-
ordinate nucleotide phosphates; however, the other
mutations, relative to the bound nucleotide phos-
phate groups, are positioned further toward the
protein surface opposite to the nucleotide bind-
g subunits meet, may constitute a region for KD
interaction, and we thus refer to it as the putative
kinase domain interaction face (Figs. 1B and 3).
The phosphate tunnel traverses the g subunit,
defining a large void that is capped on the KD-
refer to this loop, which includes residues 244 to
255, as the b flap. The region of the b flap that
charged residues and makes no contacts to the
hydrophobic core, suggesting the possibility for
structural rearrangement. The b flap appears to be
highly mobile (average B factors of 84.3 Å2in the
four independent b subunits, as compared with an
overall average B factor of 51.6 Å2for all protein
atoms) and adopts slightly different conformations
in the four independent copies of the structures
that affect AMPK activation are positioned within
the phosphate tunnel, between the terminal phos-
phate of bound AXP and the b flap. Because the
difference between the inhibitory (ATP) and acti-
vating (AMP) ligands is in the number of phos-
that affect kinase activation also lie within this
tunnel, it appears likely that this represents a critical
We have presented crystal structures that
define the core heterotrimeric architecture for
AMPKs. The S. pombe AMPK binds either
AMP or ATP at a single site, suggesting that
tion by displacing the inhibitory ligand ATP.
Nonetheless, possible binding of additional reg-
ulatory nucleotides in the context of nucleotide
mixtures or the holoenzyme complex cannot be
excluded. Although a detailed understanding of
the mechanism of AMPK regulation will require
structures of the holoenzyme, the structures pre-
sented here should provide an entry point for the
rational design of AMPK-directed therapeutics.
References and Notes
1. B. B. Kahn, T. Alquier, D. Carling, D. G. Hardie,
Cell Metab. 1, 15 (2005).
2. D. G. Hardie, Endocrinology 144, 5179 (2003).
3. B. E. Kemp et al., Biochem. Soc. Trans. 31, 162 (2003).
4. T. Leff, Biochem. Soc. Trans. 31, 224 (2003).
5. D. G. Hardie, S. A. Hawley, J. W. Scott, J. Physiol. 574, 7
6. S. B. Jorgensen, E. A. Richter, J. F. Wojtaszewski,
J. Physiol. 574, 17 (2006).
7. N. Musi, H. Yu, L. J. Goodyear, Biochem. Soc. Trans. 31,
8. Y. Minokoshi et al., Nature 415, 339 (2002).
9. T. Yamauchi et al., Nat. Med. 8, 1288 (2002).
10. R. R. Banerjee et al., Science 303, 1195 (2004).
11. V. Lumbreras, M. M. Alba, T. Kleinow, C. Koncz, M. Pages,
EMBO Rep. 2, 55 (2001).
12. O. Vincent, R. Townley, S. Kuchin, M. Carlson, Genes Dev.
15, 1104 (2001).
13. S. M. Warden et al., Biochem. J. 354, 275 (2001).
14. G. Polekhina et al., Structure 13, 1453 (2005).
15. J. Adams et al., Protein Sci. 13, 155 (2004).
16. J. Scott et al., J. Clin. Investig. 113, 274 (2004).
17. T. Daniel, D. Carling, J. Biol. Chem. 277, 51017 (2002).
18. S. Meyer, R. Dutzler, Structure 14, 299 (2006).
19. M. D. Miller et al., Proteins 57, 213 (2004).
20. R. Zhang et al., Biochemistry 38, 4691 (1999).
21. S. Meyer, S. Savaresi, I. C. Forster, R. Dutzler, Nat. Struct.
Mol. Biol. 14, 60 (2007).
22. T. J. Iseli et al., J. Biol. Chem. 280, 13395 (2005).
23. V. Nayak et al., Structure 14, 477 (2006).
24. M. J. Rudolph, G. A. Amodeo, Y. Bai, L. Tong, Biochem.
Biophys. Res. Commun. 337, 1224 (2005).
25. W. A. Hendrickson, J. R. Horton, D. M. LeMaster, EMBO J.
9, 1665 (1990).
26. Single-letter abbreviations for the amino acid residues
are as follows: A, Ala; C, Cys; D, Asp; E, Glu; F, Phe;
G, Gly; H, His; I, Ile; K, Lys; L, Leu; M, Met; N, Asn; P, Pro;
Q, Gln; R, Arg; S, Ser; T, Thr; V, Val; W, Trp; and Y, Tyr.
27. N. Tochio et al., Protein Sci. 15, 2534 (2006).
28. K. A. Wong, H. F. Lodish, J. Biol. Chem. 281, 36434 (2006).
29. X. Zhang et al., Mol. Cell 6, 1473 (2000).
30. W. A. Wilson, S. A. Hawley, D. G. Hardie, Curr. Biol. 6,
31. A. Woods et al., J. Biol. Chem. 269, 19509 (1994).
32. E. Blair et al., Hum. Mol. Genet. 10, 1215 (2001).
33. M. Arad et al., J. Clin. Investig. 109, 357 (2002).
34. M. H. Gollob et al., Circulation 104, 3030 (2001).
35. M. H. Gollob et al., New Engl. J. Med. 344, 1823 (2001).
36. Coordinates have been deposited in the Protein Data
Bank (accession number for AMP complex is 2OOX; for
ATP complex, 2OOY). Crystallographic data were acquired
at the New York Structural Biology Center beamline X4A
of the National Synchrotron Light Source, Brookhaven
National Laboratory. We thank T. Burke and G. Ahlsen for
help in biochemical experiments; X. Jin and C. Ciatto for
crystallographic help; H. Mu for technical assistance;
B. Chait, M. Cadene, and M. Gawinowicz for advice on
mass spectrometry analyses; M. Gawinowicz for help in
performing them; W. A. Hendrickson and P. D. Kwong for
critique of the manuscript; and the reviewers for excellent
suggestions. This work was supported in part by a
National Institute of Diabetes and Digestive and Kidney
Diseases grant and a Jules and Doris Stein Research to
Prevent Blindness Foundation professorship award to L.S.
Supporting Online Material
Figs. S1 to S8
13 November 2006; accepted 29 January 2007
Published online 8 February 2007;
Include this information when citing this paper.
Structure of Nup58/45 Suggests
Flexible Nuclear Pore Diameter by
Ivo Melc ˇák, André Hoelz,* Günter Blobel*
The nucleoporins Nup58 and Nup45 are part of the central transport channel of the nuclear
pore complex, which is thought to have a flexible diameter. In the crystal structure of an a-helical
region of mammalian Nup58/45, we identified distinct tetramers, each consisting of two antiparallel
hairpin dimers. The intradimeric interface is hydrophobic, whereas dimer-dimer association occurs
through large hydrophilic residues. These residues are laterally displaced in various tetramer
conformations, which suggests an intermolecular sliding by 11 angstroms. We propose that
circumferential sliding plays a role in adjusting the diameter of the central transport channel.
NPCisa ringlikestructure with an eight-foldsym-
metry. A central channel is embraced by rings and
spokes that are attached to the pore membrane
domain of the nuclear envelope (1, 2). Composed
NPC is one of the largest supramolecular as-
he nuclear pore complex (NPC) mediates
the selective exchange of macromolecules
between the nucleus and cytoplasm. The
semblies in the eukaryotic cell (~120 megadaltons
in vertebrates) (3). About 30 different nups are as-
Laboratory of Cell Biology, Howard Hughes Medical Institute,
The Rockefeller University, 1230 York Avenue, New York, NY
*To whom correspondence should be addressed. E-mail:
firstname.lastname@example.org (A.H.), email@example.com (G.B.)
VOL 31523 MARCH 2007
on April 24, 2009
as the “building blocks” of the NPC (4, 5). Mod-
eling studies have suggested that most nups are
constructed from one or two of a small number
of structural domains: coiled-coils, a-helical sole-
noids, b propellers, and natively unfolded seg-
ments containing phenylalanine-glycine repeats
(FG repeats) (6). Current translocation models
(5, 7–9) are based on the transient, low-affinity
binding of cargo-loaded transport factors to FG
repeats (10–12). Large-scale structural rear-
rangements that occur in response to transport
include the constriction and dilation of the cen-
tral channel, as well as structural alterations of
the peripheral assemblies of the NPC (2, 13–16).
However, at present, the molecular details of the
structural changes of the NPC are unknown.
The central channel of the NPC is lined by the
Nup54, and Nup45 (17). The constituents of the
Nup62 complex display a similar domain organi-
zation, in which a ~200-residue a-helical region is
flanked by FG repeats (Fig.1A) (18).Of these two
the flanking, tentacle-like FG repeats sample the
The a-helical regions of all four nups have
been predicted to form coiled-coil domains that
consist of subdomains separated by short con-
necting segments (18). Nup58 and Nup45 are
alternatively spliced forms that differ only in their
essential for cell viability (20–22) and for the
formation of a functional “minimal” NPC (23).
We assembled a Rattus norvegicus Nup62
complex that contains the predicted structured
regions of Nup62 (residues 322 to 525), Nup54
(residues 115 to 510), and Nup58 (residues 239 to
415). Through monitoring the formation and
deletion constructs was designed. We identified a
stable Nup62 complex core that is composed of
residues 322 to 525 of Nup62, residues 346 to 494
of Nup54, and residues 327 to 415 of Nup58 (Fig.
1, A and B). The minimal core domains of Nup58
and Nup45 are identical, and therefore, we refer to
Nup58 and Nup45 as Nup58/45, using the residue
numbering of Nup58 in the text below.
Reconstituted Nup62 complexes form
concentration-dependent dynamic assemblies, as
judged by size-exclusion chromatography (Fig.
1C). In addition, we analyzed the behavior of a
(fig.S1).Asdo the Nup62 complexes,the Nup58/
45 fragment exhibited substantial concentration-
dependent mobility, again indicating a dynamic,
dimer-tetramer equilibrium in solution. At low salt
assembly, which suggests that the association is
governed by electrostatic interactions (fig. S1).
From the purified Nup62 core complex, only
the Nup58/45 core domain, composed of res-
idues 327 to 415, crystallized, whereas Nup62,
Nup54, and excess Nup58/45 remained intact in
the drop solution (fig. S2). Two independent
crystallization conditions yielded two Nup58/45
crystal forms, one belonging to the tetragonal
space group P4322 and the other to the ortho-
rhombic space group C2221. The structures of
both crystal forms were solved by single-
wavelength anomalous dispersion (SAD), using
osmium-derivatized crystals. The final model
was refined to a 2.85 Å resolution with an Rwork
of 25.0% and an Rfreeof 28.8% and contains
residues 327 to 411. No electron density was
observed for the final four residues that are pre-
sumed to be disordered and therefore have been
omitted from the final model (table S1).
Nup58/45 tetramers in the two independent
crystal forms exhibit entirely different crystal
packing arrangements. The asymmetric units of
the tetragonal and orthorhombic crystals contain 4
and 18 Nup58/45 protomers, respectively. In the
packing of both crystals, the Nup58/45 tetramers
each of the two crystal forms, the tetrameric
Nup58/45 assemblies are topologically identical,
but display distinct structural differences. As the
overall findings are similar, we will focus our
discussion on the structural information obtained
from the tetragonal crystal form.
The Nup58/45 protomer folds into an antipar-
allel hairpin structure, in which two a helices (the
N and the C helix) are connected by a short loop
(Fig. 2A). The C helix protrudes from the hairpin
structure and exposes three terminal helical turns.
The four protomers in the asymmetric unit of the
tetragonal crystals are similar, and the topologi-
cally equivalent Ca carbons superimpose with a
root-mean-square deviation of ~0.4 Å. Further-
more, the protomers dimerize in an antiparallel
arrangement, in which the N and C helices of one
protomer pack against the N and C helices of the
an interprotomer angle of ~156° (Fig. 2A). The
two protomers in the dimer are primarily held
togetherby numerousvanderWaals contacts(fig.
S3), burying ~2,550 Å2of surface area. The
comparison of the average-temperature B-factors
of the entire model (~43 Å2) with the buried
dimerization interface residues (~34 Å2) suggests
a rigid and well-defined dimer interface. The
by 21 Å. Except for surface residues, the dimer is
symmetric with a pseudo–two-fold axis running
through the dimerization interface.
crystal forms consist of two dimers that interact
tion interface is composed exclusively by the four
N helices that form two antiparallel, intertwined
2) of the tetragonal crystal form have two-fold
rotational symmetry, and the two-fold symmetry
axis of each tetramer coincides with a crystallo-
graphic two-fold axis. However, although the
crystallographic two-fold axis runs perpendicular
to the long axis of the tetramerization interface in
one tetramer, it runs parallel to the long axis in the
other (Fig. 2B). The crystallographic consequence
is that two of the four protomers in each tetramer
lateral displacement of their dimer subunits along
the long axis of the tetramerization interface (Fig.
2C). We identified two similar, structurally distinct
tetrameric assemblies in the orthorhombic crystal
form (conformer 3 and 4), which exhibits an
entirely different crystal packing arrangement (fig.
S4). The conformation of these tetramers does not
appear to be dominated by the crystal packing, as
the crystal contains seven structurally independent
tetrameric assemblies that give rise to two con-
formations. All seven tetramers contain a two-fold
Fig. 1. Organization
and dynamic behavior
of the Nup62 complex.
(A) Domain structures of
Nup58, Nup45, Nup54,
regions (dark blue), un-
structured regions (light
blue), FG repeats (black
of the Nup62 complex
fragments are indicated.
The elution positions for molecular mass standards are shown. All profiles were obtained in a buffer containing 150 mM NaCl.
23 MARCH 2007VOL 315
on April 24, 2009
axis, which, as do the tetramers in the tetragonal
crystal form, run either parallel (three tetramers,
conformer 3) or perpendicular (four tetramers,
conformer 4) to the long axis of the dimer-dimer
interface. However, in contrast to the tetragonal
crystal form, only one of the seven tetrameric
assemblies is generated by a crystallographic two-
fold axis (conformer 3). Thus, the observed
tetramerization plasticity is likely an important
functional feature of the oligomerization of
Nup58/45. Furthermore, the dimer-tetramer dy-
namics observed in solution likely result from the
reversible association of Nup58/45 dimers.
In contrast to the dimerization interface, which
is almost entirely hydrophobic, the interactions
between the two dimers are solely electrostatic,
which explains the salt concentration–dependent,
dynamic behavior of Nup58/45 in solution. The
tetramerization interface is formed by an extensive
side-chain hydrogen bond network that ties the N
helices of the four protomers together (Fig. 3). In
this tetramerization interface, most of the interac-
tions occur within each pair of aligned, antiparallel
residues (Arg333, Gln344, Arg347, Gln348, Glu351,
and Asn355) of each protomer. Together, these
residues form a continuous electrostatic surface, in
side chains are distributed in the center of the
interface. The majority of the interactions are
mediated by direct intermolecular contacts, al-
though some residues make hydrogen bonds
through water molecules trapped at the protein-
protein interface. The tetragonal crystals contain
two tetrameric Nup58/45 conformers that together
contain four unique pairs of N helices. By
superposition of all four pairs of N helices, we can
~11 Å (Figs. 2C and 3). Furthermore, the
superposition reveals that the lateral displacement
between the two asymmetric tetrameric Nup58/45
conformers results from alternative hydrogen-bond
networks, in which each side chain of the same set
of interface residues has the propensity to switch
interaction partners. For example, the side chain of
Asn355engages four different residues in the two
tetramers and forms hydrogen bonds with Gln344,
Arg347, Gln348, or Glu351(Fig. 3). Overall, the
Nup58/45 core domain is highly conserved in
vertebrates (figs. S5 and S6). The residues of the
Nup58/45 that are crucially involved in the
across vertebrates, underlining the significance of
the Nup58/45 homo-tetramer.
The identification of multiple interaction states
in which rigid Nup58/45 dimers are sequentially
shifted along the dimer-dimer interface for a
distance of ~11 Å suggests an intermolecular
sliding mechanism. In general, sliding requires a
series of energetically equivalent states along a
finding that the Nup58/45 core domain forms a
salt concentration–dependent, dynamic dimer-
tetramer equilibrium in solution; this implies that
Nup58/45 dimerslooselyassociate and are indeed
capable of freely rearranging into distinct tetra-
meric conformations. Each of the four states por-
trays a snapshot of the structural alterations that
occur when two Nup58/45 dimers slide against
each other on an ~11 Å sliding pathway (Figs. 2C
and 3, and movies S1 to S3).
The rearrangement of the dimer-dimer interac-
tion surface that results in the formation of alter-
native hydrogen-bond networks is enabled by an
tural features: (i) The interaction surface is con-
checkerboard-like distribution and that are capable
of switching interaction partners by acting alterna-
or as both. For example, Asn355alternates its roles
in electrostatic interactions with Gln344, Arg347,
Gln348, and Glu351; (ii) the plasticity of the in-
teraction surface is supported by the relatively high
flexibility of long side chains. In contrast to short
side chains, long side chains are capable of
sampling a substantially larger volume for interac-
tion partners, as illustrated by the alternate con-
formations of the side chains of Arg333and Gln344
in the different tetrameric assemblies.
The central channel of the NPC has the ability
to alter its diameter (2, 13, 14, 16). The mutual
Fig. 2. Structures of tetrameric
Nup58/45 assemblies from the
tetragonal crystal form. (A) Rib-
bon representation of a Nup58/
45 dimer, showing protomer A
(right) a 90° rotated view. A
pseudo–two-fold axis (red) is
indicated. (B) Ribbon represen-
tations of the two tetrameric
conformers, indicating the loca-
tion of the crystallographic two-
fold axes (orange), which run
differently through the two tet-
ramers. (Right) A 90° rotated
view. The different coloring of
dimers (blue, conformer 1; pur-
ple, conformer 2) illustrates the
alternative tetrameric configu-
rations of the two conformers.
Symmetry-related protomers are
indicated. Conformers 1 and 2
A and B, C and D, and their
symmetry-related protomers A′
andB′andC′ and D′,respective-
ly. (C) Superposition of the two
tetrameric Nup58/45 conform-
two tetrameric assemblies are
superimposed onto protomer A
of conformer 1 to highlight the
lateral shift between the differ-
ent conformers. For clarity, only
one protomer of each superpo-
sition is colored. The N and C
termini are labeled according
to (B). The inset is expanded in
VOL 31523 MARCH 2007
on April 24, 2009
tetramer can be altered by a sliding distance of at
least ~11 Å. Nup58 and Nup45 are the most
that circumferential sliding of Nup58/45 in the
in response to transport activity. In such a poten-
tially circular, eight-fold symmetric arrangement
of tetrameric Nup58/45 modules in which each
can expand by ~11 Å, the diameter of the channel
could increase by ~30 Å (Fig. 4, A and B).
In the assembled NPC, Nup58 and Nup45 are
intimately associated with Nup54 and Nup62
share a similar domain organization and contain
conserved, amphipathic a-helical regions that may
also permit sliding (figs. S7 to S9). A multiple
beltlike arrangement of such “sliding modules” in
the channel perimeter along a nucleocytoplasmic
axis may facilitate localized changes in channel
diameter as cargo passes across (Fig. 4C). The FG
repeats are flanking the a-helical regions of the
channel nups and function as a restrictive barrier
and transport facilitator. It is conceivable that
binding of transport factors and substrate to FG
repeats is coupled to sliding of the amphipathic
helices in the perimeter of the transport channel.
The sliding of such a-helical regions may
extend to other nucleoporins,because the export
of preribosomal subunits has been shown to de-
pend on the yeast homologs of Nup58 and
Nup62 (27), as well as the coiled-coil domains
of Nup214 and Nup88 (28). Our hypothesis of
large-scale sliding facilitated by a-helical sur-
faces to adjust the diameter of a transport chan-
nel for the passage of cargo is to our knowledge
not found elsewhere and may be limited to
macromolecular transport across the NPC.
References and Notes
1. D. Stoffler et al., J. Mol. Biol. 328, 119 (2003).
2. M. Beck et al., Science 306, 1387 (2004).
3. R. Reichelt et al., J. Cell Biol. 110, 883 (1990).
4. J. M. Cronshaw, A. N. Krutchinsky, W. Zhang, B. T. Chait,
M. J. Matunis, J. Cell Biol. 158, 915 (2002).
5. M. P. Rout et al., J. Cell Biol. 148, 635 (2000).
7. I. Ben-Efraim, L. Gerace, J. Cell Biol. 152, 411 (2001).
8. K. Ribbeck, D. Gorlich, EMBO J. 20, 1320 (2001).
9. R. Y. H. Lim et al., Proc. Natl. Acad. Sci. U.S.A. 103,
10. R. Bayliss, T. Littlewood, M. Stewart, Cell 102, 99 (2000).
11. S. Fribourg, I. C. Braun, E. Izaurralde, E. Conti, Mol. Cell
8, 645 (2001).
12. T. A. Isgro, K. Schulten, Structure 13, 1869 (2005).
13. C. W. Akey, Biophys. J. 58, 341 (1990).
14. C. W. Akey, D. S. Goldfarb, J. Cell Biol. 109, 971 (1989).
15. E. Kiseleva, M. W. Goldberg, B. Daneholt, T. D. Allen,
J. Mol. Biol. 260, 304 (1996).
16. E. Kiseleva, M. W. Goldberg, T. D. Allen, C. W. Akey,
J. Cell Sci. 111, 223 (1998).
17. T. Guan et al., Mol. Biol. Cell 6, 1591 (1995).
18. T. Hu, T. Guan, L. Gerace, J. Cell Biol. 134, 589 (1996).
19. T. Hu, L. Gerace, Gene 221, 245 (1998).
20. E. C. Hurt, EMBO J. 7, 4323 (1988).
22. P. Grandi, N. Schlaich, H. Tekotte, E. C. Hurt, EMBO J. 14,
23. L. A. Strawn, T. Shen, N. Shulga, D. S. Goldfarb, S. Wente,
Nat. Cell Biol. 6, 197 (2004).
24. M. Slutsky, L. A. Mirny, Biophys. J. 87, 4021 (2004).
25. D. R. Finlay, E. Meier, P. Bradley, J. Horecka, D. J. Forbes,
J. Cell Biol. 114, 169 (1991).
26. C. Macaulay, E. Meier, D. J. Forbes, J. Biol. Chem. 270,
27. E. Hurt et al., J. Cell Biol. 144, 389 (1999).
28. R. Bernad, D. Engelsma, H. Sanderson, H. Pickersgill,
M. Fornerod, J. Biol. Chem. 281, 19378 (2006).
the initial stages of the project; L. Gerace for the gift of
cDNAs; B. Manjasetty (National Synchrotron Light Source)
and C. Ralston (Advanced Light Source) for their excellent
scientific support and help with x-ray measurements;
T. Huber for stimulating discussions; E. Coutavas, M. King,
and S. Etherton for help with editing of the manuscript. A.H.
was supported by a grant from the Leukemia and Lymphoma
Institute. The coordinates and structure factors have been
deposited in the Protein Data Bank (accession code 2OSZ).
Supporting Online Material
Materials and Methods
Figs. S1 to S9
Movies S1 to S3
29 September 2006; accepted 16 February 2007
Fig. 3. Moleculardetails
of the intermolecular
sliding of two Nup58/45
dimers.The direction and
the approximate sliding
distance are indicated by
black arrows. States I and
IV are derived from con-
former 1; states II and III
are derived from con-
former 2. The helices are
colored and numbered
according to Fig. 2, and
critical interface residues
are labeled and individu-
ally colored in all four
states according to state I
of the N-helix pairs that
tion interface are shown.
The electrostatic interac-
tions between two sliding
N helices are represented
by red dashed lines, and
water molecules are vi-
sualized as blue spheres.
Fig. 4. Model of pore dilation by
intermolecular sliding of Nup58/
45 tetramers. (A) Schematic repre-
sentation of the Nup58/45 sliding
module. The four N helices that
generate the tetramerization inter-
face (orange), the C helices (light
blue), and the C-terminal FG re-
peats (black) are indicated. The
sliding of the Nup58/45 dimer
surfaces formed by the N helices
is facilitated by an alternative
hydrogen bond network (red and
green thin lines). Because the C
helices with the attached FG re-
45 dimers against each other results in an overall extension of the Nup58/45 tetramer (green bars) that, in
turn, causes the dilation of the central channel of the NPC. (C) Localized changes in channel diameter
by a-helical sliding in response to transport of cargo (red spheres) across the central channel (green).
23 MARCH 2007VOL 315
on April 24, 2009