Structural and functional analysis of an essential
nucleoporin heterotrimer on the cytoplasmic
face of the nuclear pore complex
Kimihisa Yoshida, Hyuk-Soo Seo, Erik W. Debler, Günter Blobel1, and André Hoelz1,2
Laboratory of Cell Biology, Howard Hughes Medical Institute, The Rockefeller University, 1230 York Avenue, New York, NY 10065
Contributed by Günter Blobel, August 15, 2011 (sent for review August 1, 2011)
So far, only a few of the interactions betweenthe ≈30 nucleoporins
comprisingthe modular structure of the nuclear pore complex have
been defined at atomic resolution. Here we report the crystal struc-
ture, at 2.6 Å resolution, of a heterotrimeric complex, composed of
fragments of three cytoplasmically oriented nucleoporins of yeast:
Nup82, Nup116, and Nup159. Our data show that the Nup82 frag-
ment, representing more than the N-terminal half of the molecule,
folds into an extensively decorated, seven-bladed β-propeller that
forms the centerpiece of this heterotrimeric complex and anchors
both a C-terminal fragment of Nup116 and the C-terminal tail of
Nup159. Binding between Nup116 and Nup82 is mutually rein-
forced via two loops, one emanating from the Nup82 β-propeller
and the other one from the β-sandwich fold of Nup116, each con-
tacting binding pockets in their counterparts. The Nup82-Nup159
interaction occurs through an amphipathic α-helix of Nup159,
which is cradled in a large hydrophobic groove that is generated
from several large surface decorations of the Nup82 β-propeller.
Although Nup159 and Nup116 fragments bind to the Nup82 β-pro-
peller in close vicinity, there are no direct contacts between them,
consistent with the noncooperative binding that was detected bio-
chemically. Extensive mutagenesis delineated hot-spot residues for
these interactions. We also showed that the Nup82 β-propeller
binds to other yeast Nup116 family members, Nup145N, Nup100
and to the mammalian homolog, Nup98. Notably, each of the three
nucleoporins contains additional nuclear pore complex binding
sites, distinct from those that were defined here in the heterotri-
meric Nup82•Nup159•Nup116 complex.
assembly ∣ evolutionary conservation ∣ nucleo-cytoplasmic transport ∣
site-directed mutagenesis ∣ X-ray crystallography
refined our understanding of its architecture to a resolution in
the upper single-digit nanometer range (1). Collectively, these
studies showed that NPCs are embedded in ≈100 nm wide circu-
lar openings, resulting from a circumscribed fusion of the double
membrane of the nuclear envelope. The NPC consists of a sym-
metric central core and asymmetric filament-like structures that
project to the nucleoplasmic and cytoplasmic sides. The core of
the NPC displays a twofold axis of symmetry in the plane of the
membrane and an eightfold rotational symmetry in the nucleo-
cytoplasmic direction. As observed by cryoelectron tomography,
the NPC can undergo huge structural changes (2) that remain to
be characterized at the atomic level.
Beginning in the 1980s, biochemical and genetic analyses of
the NPC and its surrounding pore membrane domain of the
nuclear envelope has yielded an inventory consisting of ≈30 dis-
tinct NPC proteins termed nups (for nucleoporins) and of three
distinct integral membrane proteins termed poms (for pore mem-
brane proteins) that anchor the NPC. “Asymmetric” nups make
up the filament-like structures on either side of the symmetric
core and occur in at least eight copies, whereas “symmetric” nups
in the core occur in at least 16 copies (1).
ollowing the discovery of the nuclear pore complex (NPC) in
the 1950s, numerous electron microscopic studies since have
Nups consist of a combination of unstructured regions and
standard structural folds, such as β-propellers, α-helical sole-
noids, and coiled coils (1). Many of the unstructured regions are
marked by repetitive Phe-Gly motifs and were therefore dubbed
FG-repeat regions. Such regions occur in both symmetric and
asymmetric nups. Early on, FG repeats were identified as docking
sites for a collection of various transport factors that in turn re-
cognize signals on substrates for nuclear import and export (3, 4).
Several of the asymmetric nups also contain binding sites for
enzymes and other proteins involved in nucleo-cytoplasmic trans-
port. The local concentrations of these enzymes and proteins on
either side of the NPC are among the determinants for direction-
ality of transport to either the nucleus or the cytoplasm. Several
of these binding sites have already been characterized at the
atomic level. Moreover, some crystallographic analyses of the
extensive nucleoporin interactome, especially of the symmetric
core, have been reported (1).
In order to gain deeper insight into the structure of the cyto-
plasmic filament network of the yeast Saccharomyces cerevisiae
NPC, we assembled a heterotrimeric complex from fragments
comprising more than the N-terminal half of Nup82, the C-term-
inal domain of Nup116, and the C-terminal tail of Nup159. We
present the crystal structure of this heterotrimer, carried out ex-
tensive structure-guided mutagenesis to identify hot-spot residues
for these interactions, and showed that the interacting sites have
been highly conserved in evolution.
structure for each of three cytoplasmically exposed nucleoporins
of S. cerevisiae, Nup82, Nup159, and Nup116, is shown in Fig. 1A
(1, 5–12). It was previously reported that full-length Nup82 reacts
in an overlay blot with a C-terminal fragment of Nup159,
Nup1591223–1460(7). In further narrowing down the size of inter-
acting domains, we found that an N-terminal fragment of Nup82,
Nup821–452, (referred to as the Nup82 N-terminal domain,
Nup82NTD) and the C-terminal tail of Nup159 (Nup1591425–1460,
Nup159T) formed a 1∶1 complex with a molar mass of 58 kDa, as
determined by size exclusion chromatography combined with
multiangle light scattering (Fig. 1B). However, this complex
failed to crystallize, perhaps because of an unsaturated binding
site. A candidate for a missing binding partner was the C-terminal
Author contributions: K.Y., H.-S.S., E.W.D., G.B., and A.H. designed research; K.Y., H.-S.S.,
E.W.D., and A.H. performed research; K.Y., H.-S.S., E.W.D., G.B., and A.H. analyzed data;
and K.Y., H.-S.S., E.W.D., G.B., and A.H. wrote the paper.
The authors declare no conflict of interest.
Data deposition: The atomic coordinates have been deposited in the Protein Data Bank,
www.pdb.org (PDB ID code 3PBP).
1To whom correspondence may be addressed. E-mail: firstname.lastname@example.org or hoelz@
2Present address: California Institute of Technology, Division of Chemistry and Chemical
Engineering, 1200 East California Boulevard, Pasadena, CA 91125.
This article contains supporting information online at www.pnas.org/lookup/suppl/
www.pnas.org/cgi/doi/10.1073/pnas.1112846108 PNAS ∣ October 4, 2011 ∣ vol. 108 ∣ no. 40 ∣ 16571–16576
region of Nup116 that had previously been reported to interact
with N-terminal fragments of Nup82 in a yeast two-hybrid system
(13); furthermore, the solution structure of the C-terminal region
of Nup116 (Nup116967–1113, referred to as Nup116CTD) had
already been solved by NMR (14). Indeed, we were able to
assemble a soluble 1∶1∶1 Nup82NTD•Nup159T•Nup116CTDhet-
erotrimer with a molar mass of 76 kDa, as determined by size
exclusion chromatography combined with multiangle light scat-
tering (Fig. 1B), and we found that this complex crystallized.
Structure Determination of the Heterotrimer. The Nup82NTD•
Nup159T•Nup116CTDheterotrimer crystallized in space group
P1 with four trimeric complexes in the asymmetric unit. The
structure was solved by single anomalous dispersion (SAD) using
X-ray diffraction data of seleno-L-methionine-labeled crystals
andwas refined toa resolution of 2.6Åwith Rfreeand Rworkvalues
of 27.2% and 25.7%, respectively. For details of data collection
and refinement statistics, see Table 1. The four heterotrimers in
the asymmetric unit align with a root-mean square deviation of
≈0.5 Å, suggesting limited conformational flexibility. Because
there was no evidence that the heterotrimer formed higher-order
structures in solution, we focused our structural analyses on the
Individual Components of the Heterotrimer. The heterotrimer forms
an irregular structure of ≈90 Å × 60 Å × 50 Å and is shown in
different orientations in Fig. 2 A–C. The polypeptide chain
of Nup82NTDfolds into a seven-bladed β-propeller and lacks a
Velcro closure that typically intertwines blades 1 and 7 by provid-
ing a forth β-strand to the terminal seventh blade (15). Instead, a
short extension at the N terminus links blades 1 and 7 by forming
several hydrophobic contacts and hydrogen bonds with both
blades (Fig. 2 C and D). The β-propeller contains numerous
surface decorations consisting of α-helices, β-strands, and loops
indicated in the schematic ribbon model of Fig. 2D.
Nup116CTDfolds into a β-sandwich with numerous protruding
loops and is flanked at each end by an α-helix, termed αA and αB
(Fig. S1). One of the two β-sheets is formed by six antiparallel
β-strands, while the opposing β-sheet contains only two antipar-
allel β-strands, resulting in a groove on the molecular surface
between helix αB andstrand β5(Figs.S1 andS2).This β-sandwich
fold was first reported in the crystal structures of human
Nup98721–863and of S. cerevisiae Nup145N443–605(16, 17), as well
as in the NMR structure of Nup116CTD(14). A prominent feature
of Nup116CTDis the 19-residue loop between β6 and αB that is
critical for the interaction with Nup82NTD. Finally, Nup159T
forms a 26-residue amphipathic α-helix that binds in a surface
groove at the lower edge of the Nup82 β-propeller (Fig. 2).
Heterotrimer Interfaces. The Nup82-Nup116 interface is bipartite.
Two loops, one emanating from the Nup82 β-propeller and one
from the Nup116 β-sandwich, mediate the interaction with dis-
tinct pockets on their counterparts. One of these loops, the 3D4A
loop of Nup82, binds with a Phe-Gly-Leu (FGL) motif located
at its tip to the prominent hydrophobic groove on the Nup116
surface (Fig. 3A, Fig. S2); we therefore refer to this loop as
the “FGL loop.” The other loop, the β6-αB loop of Nup116, binds
with a distinct lysine residue to a Nup82 pocket, located at the
side of the β-propeller (Fig. 3B, Fig. S3); we therefore refer to this
loop as the “K-loop.” A salient feature of this interaction is a salt
bridge between the invariant lysine residue of Nup116 (K1063)
and an aspartate residue at the bottom of a Nup82 pocket
(D204). This electrostatic interaction is reinforced by a hydro-
phobic bracelet formed by I287, F290, and Y295, located in the
noncanonical insertion in the 4D5A connection of Nup82, and
wrapping around the apolar base of the side chain of K1063
(Fig. 3B, Fig. S3). The two sites are adjacent and form an inter-
face with a combined buried surface area of ≈1;740 Å2(Figs. S2
The Nup82-Nup159 interface is characterized by the amphi-
pathic α-helix of Nup159Tthat is situated in a large groove of
the Nup82 β-propeller, created by the exposed strands of blade
5 and the 4CD and 6CD helical insertions (Figs. 2D and 3C).
Altogether, a combined surface area of ≈1;700 Å2is buried
main organization of yeast Nup82, Nup116, and Nup159 (1). Bars denote the
fragments that were used for heterotrimer assembly and crystallization:
Nup82 N-terminal domain (NTD, blue); Nup116 C-terminal domain (CTD,
green); and Nup159 C-terminal tail (T, red). GLEBS, binding site for the RNA
export factor Gle2; DID, dynein light chain interacting domain. (B) Analysis by
size exclusion chromatography coupled to multiangle light scattering. The
differential refractive indices of Nup116CTD(red), Nup82NTD•Nup159T(blue),
and Nup82NTD•Nup159T•Nup116CTD(green) are plotted against the elution
volumes from a Superdex 200 10∕300 GL gel filtration column (GE Health-
care) with dots indicating molar masses.
Assembly of a Nup82NTD•Nup159T•Nup116CTDheterotrimer. (A) Do-
Table 1. Crystallographic analysis
a, b, c (Å)
a ¼ 61.5,
b ¼ 96.8,
c ¼ 144.3
α ¼ 106.0,
β ¼ 94.0,
γ ¼ 108.2
0.9794 (Se Peak)
α, β, γ (°)
No. of atoms
Bond angles (°)
Bond lengths (Å)
Most favored (%)
Additionally allowed (%)
Generously allowed (%)
14,690 (8.0 %)
*APS, Advanced Photon Source, Argonne National Laboratory.
†Highest-resolution shell is shown in parentheses.
‡As determined by Procheck (28).
www.pnas.org/cgi/doi/10.1073/pnas.1112846108Yoshida et al.
between the two proteins, involving numerous, primarily hydro-
phobic residues (Figs. S3 and S4).
Paralogs and Orthologs of Nup116. While yeast Nup116 has two
paralogs, Nup100 and Nup145N, only one ortholog exists in hu-
mans (hNup98). Both Nup145N and hNup98 are auto-proteolytic
cleavage products of larger polypeptide chains, encompassing
the N-terminal Nup145N or hNup98, followed by Nup145C or
hNup96, respectively (18–20). Notably, hNup98 is also synthe-
sized from an alternatively spliced mRNA that encodes for a
β-propeller with various noncanonical insertions highlighted in yellow (3D4A or “FGL” loop), gray (4CD), and orange (6CD); Nup116CTD(green) is a β-sandwich
with a β6-αB connector (“K-loop”) indicated in magenta; Nup159T(red) folds into an amphipathic α-helix. Dotted lines represent disordered regions.
(D) Schematic representation of the β-propeller of Nup82NTD. Prominent insertions and secondary structure elements are labeled. The asterisk denotes
the N-terminal region that fits into the crevice between blades 1 and 7, replacing the canonical Velcro closure observed in most β-propellers.
Structural overview of the Nup82NTD•Nup159T•Nup116CTDcomplex. (A–C) Ribbon representation from various angles. Nup82NTD(blue) folds into a
the 3D4A “FGL loop” of Nup82NTD, which binds to a groove of Nup116CTDbetween helix αB and strand β5. (B) The other site is formed by the “K-loop” of
Nup116CTD, K1063 of which binds to a Nup82 pocket with an aspartate (D204) at its bottom and a hydrophobic bracelet at its entry. (C) Nup159Tbinds to a
groove in Nup82NTDthat is formed by the 4CD and 6CD insertions and blade 5. Color code as in Fig. 2.
Close-up view of the Nup82-Nup116 and Nup82-Nup159 interfaces. Nup82NTDinteracts with Nup116CTDvia two adjacent sites. (A) One site is formed by
Yoshida et al.PNAS
October 4, 2011
C-terminal 6-kDa protein instead of hNup96 (19). Auto-proteo-
lytic cleavage at HF ↓ SKYGL requires a catalytic serine (18)that
is not conserved in the Nup116 and Nup100 paralogs (Fig. S5).
The CTDs of Nup116 and hNup98 superpose with a root-mean
square deviation of ≈2.0 Å. Strikingly, the binding of the FGL
loop of Nup82NTDto the groove of Nup116CTDmimics the inter-
action of the Tyr-Gly-Leu (YGL) portion of the auto-proteolytic
cleavage motif HF ↓ SKYGL bound to hNup98 (Fig. S1 A and D).
To directly test the mimicry of these interactions, we utilized a
construct (19) encoding hNup98CTDand a C-terminal 6-kDa pep-
tide (Fig. 4A), yielding an hNup98CTD•6-kDa heterodimer com-
posed of the auto-proteolytic cleavage products (Fig. 4 B and C).
When this human heterodimer was mixed with the yeast
pair,a chimeric Nup82NTD•Nup159T•
hNup98CTDheterotrimer was formed displacing and releasing the
6-kDa peptide (Fig. 4 B and C). We showed that the other two
Nup116 paralogs, Nup100 and Nup145N, also form heterotrimeric
complexes with Nup82NTD•Nup159T(Fig. 5 A and B).
Mutational Analyses. To identify residues critical for mediating
Nup82NTD•Nup159T•Nup116CTDheterotrimer formation, we
employed structure-guided mutagenesis and probed the mutants
for complex formation by size exclusion chromatography (Table 2).
For the interaction between Nup116CTDand the Nup82NTD•
Nup159Tpair, we identified K1063 in the K-loop of Nup116CTD
as a hot-spot residue that, when mutated to alanine, abolished
complex formation.Bycontrast, mutating D204,which formsa salt
bridge with K1063, to alanine had only a mild effect on complex
formation. However, the introduction of additional mutations,
F290A and Y295A, at the entry of the hydrophobic pocket dis-
rupted complex formation (Figs. S3 and S4). We refer to this
mutant as Nup82DFY. As measured by isothermal titration calori-
metry (ITC), the deletion of the FGL loop in Nup82NTDmerely
resulted in fivefold reduction in binding affinity to Nup116CTD
(Fig. S6 A and B), thus rendering the K-loop of Nup116CTDthe
principal determinant for the Nup82NTD-Nup116CTDinteraction.
To probe the Nup82NTD-Nup159Tinteraction, we mutated five
large hydrophobic residues located in the 6CD insertion of the
Nup82 β-propeller to alanine (Figs. S3 and S4). While the indi-
vidual mutations of L393, I397, L402, L405, and F410 had no
detectable effect (Table 2), their combination abolished complex
formation. We refer to this mutant as Nup82LILLF. As expected
from the crystal structure, binding of Nup82LILLFto Nup116CTD
analysis between Nup82NTD• Nup159Tand the yeast Nup116CTDhomologs
(A) Nup100CTDand (B) Nup145NCTD. Size exclusion chromatography analysis
of Nup82NTD•Nup159T(blue), Nup100CTD(red) and Nup145NCTD(red), and
their mixtures (green).
Binding of yeast paralogs of Nup116CTDto Nup82NTD. Interaction
of human Nup98 (1, 18) indicating the C-terminal domain (CTD, green) and
the auto-proteolytic 6-kDa protein (gray), as well as the auto-proteolytic
cleavage site (red arrow). (B) Size exclusion chromatography analysis of
purified yeast Nup82NTD•Nup159T(blue), human Nup98CTD•6-kDa (red) and
a mixture (green) in an ≈1∶2 ratio. Note formation of a trimeric Nup82NTD•
Nup159T•hNup98CTDcomplex leads to displacement of the 6-kDa protein
(inset). (C) Fractions, indicated by gray bars in (B), of hNup98CTD•6-kDa
(top), Nup82NTD•Nup159T(center), and their mixture (bottom) were ana-
lyzed by SDS-PAGE; proteins are indicated on the right and molecular weight
standards on the left.
Human Nup98CTDbinds to yeast Nup82NTD. (A) Domain organization
www.pnas.org/cgi/doi/10.1073/pnas.1112846108Yoshida et al.
was not affected and, vice versa, Nup82DFYretained binding
to Nup159T. The combination of these mutations in the
Nup82DFY-LILLFvariant resulted in loss of binding to both
Nup116CTDand Nup159T. These results are in accord with ITC
data that revealed comparable affinities of Nup82NTDfor
Nup116CTDin the presence and absence of Nup159T[dissocia-
tion constants (Kd) of ≈50 and ≈60 nM, respectively] (Fig. S6 A
and C). We conclude that Nup82NTDfacilitates noncooperative
binding to Nup159Tand Nup116CTD.
Based on our result that the CTDs of the Nup116 family
Nup82NTD•Nup159T, we tested whether the identified mutations
would also abolish complex formation with these proteins.
Indeed, as with Nup116CTD, complex formation is prevented
by the Nup82DFYmutant, as well as by the corresponding K-loop
mutations, Nup100K910A, Nup145NK558A, and hNup98K814A
and hNup98 interactwith
In Vivo Analyses. In yeast, the deletion of Nup82 and Nup159 is
lethal (6, 9–11), whereas that of Nup116 yields a temperature-
sensitive phenotype (22). To test the physiological relevance of
the interfaces that we have identified here, we analyzed haploid
S. cerevisiae strains from which each of the three genes were de-
leted and replaced by various GFP-tagged wild-type and mutant
constructs. The strains were then analyzed for growth and nuclear
rim staining as an indicator for NPC incorporation.
For Nup82, deletion of its NTD, Nup82ΔNTD, yields growth de-
fects with increasing temperatures (Fig. 6 A and B). While neither
the Nup82DFYnor the Nup82LILLFmutants affected cell growth,
their combination in Nup82DFY-LILLFmimicked the growth pheno-
type of the NTD deletion (Fig. 6 A and B). Likewise, detectable
nuclear rim staining was observed only for the Nup82DFYor the
Nup82ΔNTDmutants (Fig. 6C).
For Nup116 as well as for Nup159, the deletion of the
crystallized domains from full-length proteins did not yield any
detectable defect in either growth or nuclear rim localization
(Figs. S7 and S8), indicating that these nucleoporins contain
mutants, but not for the Nup82DFY-LILLF
GFP-labeled Nup82 constructs, colored as in Fig. 1A. The black and red arrows
indicate the positions of the DFY and LILLF mutations, respectively. (B) Yeast
growth analysis using a nup82Δ strain transformed with the indicated
GFP-Nup82 constructs. 10-fold serial dilutions were spotted on SD-Leu plates
and grown for 2–3 d at the indicated temperatures. The combination of
the DFY and LILLF mutations in one mutant (Nup82DFY–LILLF) has the same
effect on growth as the deletion of the entire NTD (Nup82ΔNTD). (C) In vivo
localization of the GFP-Nup82 constructs at 37°C. Both the Nup82DFY–LILLFand
the Nup82ΔNTDmutants fail to localize to the nuclear rim. The scale bars
represent 5 μm.
In vivo analysis of Nup82 mutants. (A) Domain organization of the
Table 2. Biochemical interaction analysis
Mutations *Binding partner MutationRelative binding†
*Nup82NTDcarries the C396S mutation, referred to as wild-type (wt).
†Relative binding was estimated in size exclusion chromatography from wild type-like binding (+++) to no detectable binding (−).
‡Nup159Tbinding experiments were carried out with a His6-SUMO-Nup159Tfusion protein.
Yoshida et al. PNAS
October 4, 2011
additional domains for targeting to the NPC. In fact, for Nup116, Download full-text
we identified an additional upstream region (residues 686–967)
that is involved in NPC targeting (Fig. S7C).
We found that an N-terminal fragment comprising more than half
of the Nup82 molecule folds into a β-propeller that serves as the
binding platform for the C-terminal domain of Nup116 and the
C-terminal tail of Nup159. For each of these interactions in the
Nup82NTD•Nup159T•Nup116CTDheterotrimer, we identified
hot-spot residues by mutational analyses and verified their critical
nature by biochemical, biophysical, and in vivo experiments.
Equivalent domains of the yeast Nup116 paralogs, Nup100
and Nup145N, and the mammalian homolog, hNup98, bound
to the Nup82 β-propeller in an evolutionarily conserved fashion
and in a mutually exclusive manner. Of these, only Nup116 and
hNup98 possess a binding site for Gle2 (or Rae1 or mrnp41) that
functions in RNA export (5, 21–23). Hence, binding of either
Nup116, Nup145N, or Nup100 could affect mRNA export. It is
conceivable, for example, that replacement of yeast Nup116 by
Nup100 or Nup145N in the M or S phase of the cell cycle, or
during stationary growth, could coordinate reduced transcription
with a slowdown of mRNA export.
Our in vivo targeting data show that the interaction sites in the
heterotrimeric complex are not the only regions by which these
nucleoporins are anchored to the NPC. For example, although
the C-terminal domain of Nup116 yielded some nuclear rim
staining, consistent with targeting to the NPC, we identified an
additional targeting site upstream of the crystallized C-terminal
domain. Likewise, after removal of the crystallized C-terminal tail
of Nup159, the tailless Nup159 still displayed rim staining and
showed only a slight accumulation in the cytoplasm, whereas a
larger C-terminal truncation failed to target to the nuclear rim,
as previously reported (8). For Nup82, the simultaneous muta-
tional disruption of the two binding sites to Nup116 and Nup159
resulted in cytoplasmic localization with no detectable nuclear
rim staining. However, rim staining was detected when only one
of the binding sites was disrupted, suggesting that the loss of a
single interaction can be tolerated. In addition to the two inter-
actions described here, the C-terminal α-helical domain of Nup82
has previously been reported to interact with the α-helical region
of the nucleoporin Nsp1 that is part of the transport channel in
the core of the NPC (1, 24).
Taken together, our data indicate that Nup82 is a central in-
teraction platform for the simultaneous binding of at least three
nucleoporins, Nup116 (or their paralogs, Nup100 and Nup145N),
Nup159, and Nsp1. The elucidation of the interactions in the
Nup82NTD•Nup159T•Nup116CTDheterotrimer on the cytoplas-
mic face of the NPC is another substantial step toward a gradual
reconstruction of the NPC at atomic resolution.
The details of protein expression, purification, crystallization, structure deter-
mination, protein interaction analysis, multiangle light scattering, and in vivo
experiments are described in the SI Text published online. In short, DNA
fragments of S. cerevisiae Nup82, Nup159, Nup100, Nup145N, Nup116, and
human Nup98 were amplified by PCR and cloned into the vectors pET21a
(Novagen), pETDuet-1 (Novagen), pET28a that was modified to contain an
N-terminal PreScission protease cleavable His6-tag, and pET28b that was
modified to contain an N-terminal His6-SUMO tag (25, 26). Point mutants
were generated by QuikChange site-directed mutagenesis (Stratagene)
and confirmed by DNA sequencing. Details of the bacterial expression con-
structs are listed in Table S1. All proteins were expressed in Escherichia coli
using the appropriate expression constructs and purified using several
chromatographic techniques. X-ray diffraction data were collected at the
National Institute of General Medical Sciences and National Cancer Institute
Collaborative Access Team (GM/CA-CAT) beamline at the Advanced Photon
Source (APS), Argonne National Laboratory (ANL). The 5 μm “minibeam”
collimator setup (27) was critical for obtaining excellent X-ray diffraction
data from the twinned crystals. The structure was solved by SAD, using data
obtained from selenomethionine-labeled crystals. Data collection and refine-
ment statistics are summarized in Table 1.
ACKNOWLEDGMENTS. We thank Andrew Davenport for technical support;
Alina Patke for discussions; David King (HHMI, UC Berkeley) for mass spectro-
metry analysis; Michael Becker, Robert Fischetti, Craig Ogata, and Ruslan
Sanishvili (APS) for beamline support; the Biophysics Core Facility of the
University of Colorado, Denver, for isothermal titration calorimetry. E.W.D.
was supported by a Dale F. and Betty Ann Frey Fellowship of the Damon
Runyon Cancer Research Foundation, DRG-1977-08, and A.H. by a SCOR grant
from the Leukemia and Lymphoma Society.
1. Hoelz A, Debler EW, Blobel G (2011) The structure of the nuclear pore complex. Annu
Rev Biochem 80:613–643.
2. Beck M, Lucic V, Forster F, Baumeister W, Medalia O (2007) Snapshots of nuclear pore
complexes in action captured by cryo-electron tomography. Nature 449:611–615.
3. Radu A, Moore MS, Blobel G (1995) The peptide repeat domain of nucleoporin Nup98
functions as a docking site in transport across the nuclear pore complex. Cell
4. Cook A, Bono F, Jinek M, Conti E (2007) Structural biology of nucleocytoplasmic
transport. Annu Rev Biochem 76:647–671.
5. Wente SR, Rout MP, Blobel G (1992) A new family of yeast nuclear pore complex
proteins. J Cell Biol 119:705–723.
6. Hurwitz ME, Blobel G (1995) NUP82 is an essential yeast nucleoporin required for poly
(A)+ RNA export. J Cell Biol 130:1275–1281.
7. Hurwitz ME, Strambio-de-Castillia C, Blobel G (1998) Two yeast nuclear pore complex
proteins involved in mRNA export form a cytoplasmically oriented subcomplex. Proc
Natl Acad Sci USA 95:11241–11245.
8. Del Priore V, et al. (1997) A structure/function analysis of Rat7p/Nup159p, an essential
nucleoporin of Saccharomyces cerevisiae. J Cell Sci 110:2987–2999.
9. Grandi P, et al. (1995) A novel nuclear pore protein Nup82p which specifically binds to
a fraction of Nsp1p. J Cell Biol 130:1263–1273.
10. Gorsch LC, Dockendorff TC, Cole CN (1995) A conditional allele of the novel repeat-
containing yeast nucleoporin RAT7/NUP159 causes both rapid cessation of mRNA
export and reversible clustering of nuclear pore complexes. J Cell Biol 129:939–955.
11. Kraemer DM, Strambio-de-Castillia C, Blobel G, Rout MP (1995) The essential yeast
nucleoporin NUP159 is located on the cytoplasmic side of the nuclear pore complex
and serves in karyopherin-mediated binding of transport substrate. J Biol Chem
12. Stelter P, et al. (2007) Molecular basis for the functional interaction of dynein light
chain with the nuclear-pore complex. Nat Cell Biol 9:788–796.
13. Ho AK, et al. (2000) Assembly and preferential localization of Nup116p on the cyto-
plasmic face of the nuclear pore complex by interaction with Nup82p. Mol Cell Biol
14. Robinson MA, et al. (2005) Multiple conformations in the ligand-binding site of the
yeast nuclear pore-targeting domain of Nup116p. J Biol Chem 280:35723–35732.
15. Paoli M (2001) Protein folds propelled by diversity. Prog Biophys Mol Biol 76:103–130.
16. Hodel AE, et al. (2002) The three-dimensional structure of the autoproteolytic, nuclear
pore-targeting domain of the human nucleoporin Nup98. Mol Cell 10:347–358.
17. Sampathkumar P, et al. (2010) Structures of the autoproteolytic domain from the
Saccharomyces cerevisiae nuclear pore complex component, Nup145. Proteins
18. Rosenblum JS, Blobel G (1999) Autoproteolysis in nucleoporin biogenesis. Proc Natl
Acad Sci USA 96:11370–11375.
19. Fontoura BM, Blobel G, Matunis MJ (1999) A conserved biogenesis pathway for nu-
cleoporins: proteolytic processing of a 186-kilodalton precursor generates Nup98
and the novel nucleoporin, Nup96. J Cell Biol 144:1097–1112.
20. Teixeira MT, et al. (1997) Two functionally distinct domains generated by in vivo clea-
vage ofNup145p:a novelbiogenesispathway fornucleoporins.EMBO J16:5086–5097.
21. Bailer SM, et al. (1998) Nup116p and nup100p are interchangeable through a con-
served motif which constitutes a docking site for the mRNA transport factor gle2p.
EMBO J 17:1107–1119.
22. Wente S, Blobel G (1993) A temperature-sensitive NUP116 null mutant forms a nuclear
envelope seal over the yeast nuclear pore complex thereby blocking nucleocytoplas-
mic traffic. J Cell Biol 123:275–284.
23. Ren Y, Seo HS, Blobel G, Hoelz A (2010) Structural and functional analysis of the inter-
action between the nucleoporin Nup98 and the mRNA export factor Rae1. Proc Natl
Acad Sci USA 107:10406–10411.
24. Bailer SM, Balduf C, Hurt E (2001) The Nsp1p carboxy-terminal domain is organized
into functionally distinct coiled-coil regions required for assembly of nucleoporin
subcomplexes and nucleocytoplasmic transport. Mol Cell Biol 21:7944–7955.
25. Hoelz A, Nairn AC, Kuriyan J (2003) Crystal structure of a tetradecameric assembly
of the association domain of Ca2+/calmodulin-dependent kinase II. Mol Cell
26. Mossessova E, Lima CD (2000) Ulp1-SUMO crystal structure and genetic analysis reveal
conserved interactions and a regulatory element essential for cell growth in yeast.
Mol Cell 5:865–876.
27. Fischetti RF, et al. (2009) Mini-beam collimator enables microcrystallography experi-
ments on standard beamlines. J Synchrotron Radiat 16:217–225.
28. Laskowski RA, MacArthur MW, Moss DS, Thornton JM (1993) PROCHECK—a program
to check the stereochemical quality of protein structures. J App Crystallogr
www.pnas.org/cgi/doi/10.1073/pnas.1112846108Yoshida et al.