(A) Domain organization of Pom152^FL and four truncations drawn to scale. Pom152^FL exhibits a three-domain organization with the domain inside the NPC (NPC), the transmembrane segment (TM), and the domain inside the nuclear envelope lumen (luminal domain). The numbers indicate amino acid residue positions. Horizontal grey lines for the truncations represent the number of amino acid residues in each segment.
(B) 35 representative negative-stain EM class averages of Pom152^FL. The number of particles in each class is shown. Bar, 10 nm.
(C) Representative random conical tilt 3D maps (right) are aligned to each of the corresponding class averages (left). The number of particles in each class is shown. Bars, 10 nm.
(D) Negative-stain EM density map of Pom152^FL. The nine globular domains in C-terminal lumen (1 to 9) and the N-terminal head region containing the NPC-associating and TM domains are indicated. Bar, 50 Å.
(E) The average inter-domain angle for the last 7 repetitive regions was estimated in 37 representative Pom152FL class averages using the ImageJ angle tool. One angle was measured between domains 3 and 9, its difference from 180° was determined and the resulting value divided between the seven globular domains involved in the estimation. Distribution of the resulting inter-domain angles is shown in a Kernel density plot with a peak of −4.1° indicating a small negative curvature for the particles.
(F) Representative negative-stain EM class averages of three assigned views of Pom152^FL, Pom152^1–1135, and Pom152^1–936. The number of particles in each class is shown. Bar, 10 nm.
(G) End-to-end distances of Pom152^FL, Pom152^1–1135, and Pom152^1–936 in samples of 30, 28, and 33 class averages, respectively. The lines on the data points indicate the mean and SD: 38.7 ±1.5 nm, 28.9 ±1.6 nm and 24.4 ±1.0 nm for the samples, respectively.
See also and .

(A) Growth phenotypes of Pom152^FL tagged with Protein A or prescission protease cleavage site (PPX)-Protein A and related truncation mutants (see ). Serial 10-fold dilutions of cells were spotted on YPD plates in the absence or presence of 0.3% and 0.4% benzyl-OH at 30°C or on YPD plates an d grown at the indicated temperatures for 1–3 days.
(B) Subcellular localization of benzyl-OH treated Pom152-GFP constructs. Panels show the localization of the indicated genomically tagged Pom152-GFP or Nup188-mCherry (used as NPC localization control) constructs as determined by fluorescence microscopy. Cells were grown on liquid yeast minimal media supplemented with 2% glucose at 30° C (untreated) or with an additional 0.1% benzyl-OH for 3 h at 30° C (benzyl-OH). Differential interference contrast (DIC). Bar, 5 microns.
See also and .

(A) Superimposition of 20 lowest-energy structures of Pom152^718–820 that satisfy the NMR restraints best. The two β-sheets are made of the ABE (blue) and C′CFGG′A′ β-strands (cyan).
(B) Topology of the Pom152^718–820 fold. β-strands are represented as thick arrows and the linkers connecting them as lines. Each strand is named following the standard Ig-like domain annotation (). Scale bar, 5 Å.
(C and D) Two views of the best-scoring structure, showing the arrangement of β-strands and features labeled as in panel B.
(E and F) Sequence conservation of Pom152^718–820 among 37 fungal homologs (), plotted on the surface of the Pom152^718–820 structure; low conservation in white, high conservation in red.
(G and H) Electrostatic potential on the surface of the Pom152^718–820 structure, calculated using the PyMOL tool APBS; negative (−2 kT/e) and positive (+2 kT/e) potentials are shown in red and blue, respectively.
See also and .

(A) Domain organization of Pom152^FL and its domains drawn to scale. The head region contains the domain that faces the NPC inner ring (NPC) and the transmembrane segment (TM), followed by the lumenal domain composed of nine Ig-like repeats. The amino acid boundaries for each Ig-like repeat are indicated below them. The repeat analyzed by NMR (Ig-4, Pom152^718–820) is highlighted in light blue.
(B) Comparative models of the remaining 8 Ig-like domains were built using the Pom152^718–820 NMR structure as the template, followed by superposing them on the Pom152^718–820 structure. The 8 domains share statistically significant sequence alignments to Pom152^718–820 (E-values ranging from 3.3×10⁻⁴⁹ to 6.8×10⁻³⁹).
(C) Sequence conservation among the 9 Ig-like domains was mapped on the structure of Pom152^718–820 using ConSurf (). Low conservation, white; high conservation, orange.
(D) Domain organization of human Nup210. See legend of .
(E) Superposition of comparative models for yeast Pom152^718–820 (orange) and human Nup210^1079–1152 (blue).
See also .

The integrative structure determination of Pom152^FL proceeds through 4 stages: (1) gathering data, (2) representing and translating data into spatial restraints, (3) conformational sampling to produce an ensemble of structures that satisfies the restraints, and (4) analyzing, assessing, and validating the ensemble structures. The modeling protocol (i.e., stages 2, 3, and 4) was scripted using the Python Modeling Interface (PMI), version 4d97507, a library for modeling macromolecular complexes based on our open-source Integrative Modeling Platform (IMP) package, version 2.6 (http://integrativemodeling.org) ().

(A) (left) The localization probability density map computed from 364 superposed structures that satisfy the input spatial restraints shows the location of each of the 9 Ig-like domains (ranging from blue to red). The negative-stain EM density map is superposed in grey. A representative molecular model of Pom152^LD (ribbon plot) was obtained by adjusting the relative orientations of adjacent Ig-like domains to resemble those in known cadherin structures (PDB codes 1L3W, 1EPF, 1NCI, 4ZI9, and 5K8R) (; ; ; ; ).
(right) Validation of the integrative structure of Pom152^LD by SAXS data for five Pom152 segments, spanning residues 718–820 (Ig-4), 718–920 (Ig-4,5), 603–820 (Ig-3,4), 919–1020 (Ig-6), and 718–1148 (Ig-4,5,6,7). The shapes of these segments in our integrative structure match the envelopes (ab initio shapes) computed from the corresponding SAXS profiles.
(B) Fit of 16 copies of Pom152^LD into the yeast NPC map (). A good fit positions two copies of the extended Pom152^LD molecule in an anti-parallel fashion on top of each other, forming a homodimer; we only show the potential arrangement of the antiparallel homodimer, which is implied by the C₂-symmetry of the NPC (; ; ).
See also , and .