Nuclear Pore Complex

The Nuclear Pore Complex (NPC) is the largest macromolecular protein assembly in the eukaryotic cell—a ~110 to 120 Megadalton (MDa) cylinder with 8-fold rotational symmetry (~120 nm outer diameter, ~40 to 50 nm central transport channel diameter) that perforates the double lipid bilayer of the nuclear envelope at sites where the outer nuclear membrane (ONM, continuous with the rough endoplasmic reticulum) fuses with the inner nuclear membrane (INM). Because eukaryotic transcription and pre-mRNA splicing occur inside the nucleus while mRNA translation by ribosomes occurs in the cytosol, every proliferating human cell maintains 2,000 to 5,000 nuclear pore complexes, each capable of translocating up to 1,000 macromolecular cargoes per second (~80 MDa·s⁻¹ of mass flux per pore) in both directions simultaneously.

Remarkably, the NPC operates as a dual-mode molecular sieve and active phase-partitioning gate without possessing any motor ATPase or GTPase subunits within its structural scaffold:

1. Passive Size-Exclusion Diffusion (< 30–40 kDa): Water, inorganic ions, nucleotides, and small globular proteins with hydrodynamic radii < 2.5–3.0 nm (molecular weight < 30 to 40 kDa, such as histones or small metabolites, though histones are still actively imported in vivo) diffuse passively across the pore in milliseconds.
2. Facilitated Karyopherin-Mediated Transport (> 40 kDa up to 40 MDa): Large proteins (such as DNA/RNA polymerases, nuclear hormone receptors, and transcription factors), ribosomal 40S and 60S subunits (~1.4–2.8 MDa), and intact viral capsids cannot penetrate the central channel on their own. Instead, they display specific Short Linear Motifs (SLiMs)—either a basic Nuclear Localization Signal (NLS) for import or a leucine-rich Nuclear Export Signal (NES) for export—that are recognized by soluble transport receptors of the Karyopherin-β (Kap-β / Importin-β and Exportin) superfamily, which rapidly partition through the central Phenylalanine–Glycine (FG)-repeat nucleoporin barrier and derive directional energy from a steep nucleocytoplasmic gradient of the small Ras-family GTPase Ran (Ran-GTP vs. Ran-GDP).

Solved at near-atomic composite resolution (~3.5 to 6.0 Å) by integrating cryo-electron tomography (cryo-ET), X-ray crystallography, cross-linking mass spectrometry, and AlphaFold2 structure prediction (by Martin Beck, André Hoelz, Michael Rout, Patrick Cramer, Yigong Shi, and colleagues), each vertebrate NPC is constructed from ~1,000 individual protein chains representing multiple copies (8, 16, 32, or 48 copies per pore) of only ~30 distinct genes called nucleoporins (Nups), organized into four concentric architectural shells:

1. Luminal Transmembrane Ring (POMs — Pore Membrane Proteins): Three integral membrane proteins (NDC1, POM121, and GP210 in vertebrates) span the curved pore membrane where the inner and outer nuclear membranes fuse, anchoring the entire soluble NPC scaffold to the lipid bilayer.
2. Outer Coaxial Rings (Cytoplasmic Ring and Nuclear Ring — The Y-Complex [Nup107–160]): Flanking the central equator on both the cytoplasmic and nucleoplasmic faces are two giant octagonal rings constructed from 16 copies each (32 total per NPC, arranged in two concentric head-to-tail octameric rings per face in humans) of the 10-subunit Y-complex (Nup107–160 subcomplex; yeast Nup84 complex). Crucially, the structural subunits of the Y-complex (Nup133, Nup107, Nup96, Nup85, Sec13, Seh1) fold into 7-bladed WD40 β-propellers fused to α-solenoid rods—the exact same protocoatomer architecture found in COPII (Sec13/Sec31), COPI, and clathrin vesicle coats, providing direct structural evidence that the eukaryotic nuclear envelope and NPC co-evolved with the endomembrane vesicle-coating machinery from a common Asgard archaeal ancestor during eukaryogenesis.
3. Inner Ring (Spoke) Complex (Nup205, Nup188, Nup155, Nup93, Nup62–58–54): Sandwiched equatorially between the cytoplasmic and nuclear outer rings, the Inner Ring forms the central 40–50 nm waist of the pore and anchors the dense brush/hydrogel of FG-nucleoporins that project into the central transport lumen.
4. Asymmetric Cytoplasmic Filaments (Nup358/RanBP2, Nup214, Gle1) vs. Nuclear Basket (Tpr, Nup153, Nup50): Whereas the core scaffold is symmetric across the membrane plane, the two entrances are strictly asymmetric to enforce directional cargo capture and disassembly:
   • Cytoplasmic Filaments: Eight flexible arms dominated by the giant 358 kDa nucleoporin Nup358 (RanBP2) project into the cytosol, carrying four Ran-binding domains, a SUMO E3 ligase domain, and a high-affinity docking site for RanGAP1 (Ran GTPase-Activating Protein 1), alongside Nup214/Nup88 and the ATP-dependent DEAD-box RNA helicase DDX19 (Dbp5) + Gle1 that strip export factors off emerging mRNAs.
   • Nuclear Basket: Eight protein filaments constructed from the coiled-coil dimer Tpr (Translocated Promoter Region), Nup153, and Nup50 converge into a distal ring (~50 nm into the nucleoplasm) that docks the TREX-2 transcription–export complex and promotes rapid Ran-GTP-triggered release of imported cargoes.

How can a single 40 nm-wide channel simultaneously exclude an inert 60 kDa bovine serum albumin molecule while allowing a 1,000 kDa Importin-β–cargo complex to diffuse across in 5 to 10 milliseconds?

Lining the inner ring of the NPC are over 200 intrinsically disordered polypeptide chains (Nup98, Nup62, Nup54, Nup58, Nup153, Nup214) that contain thousands of Phenylalanine–Glycine (FG) dipeptide repeats (GLFG, FXFG, and FG motifs) separated by polar, uncharged spacer sequences (Ser, Thr, Asn, Gln). As demonstrated by Dirk Görlich, Rout, Lim, and colleagues, in the crowded center of the pore (local FG repeat concentration ~30 to 50 mM), aromatic phenylalanine (Phe) rings from neighboring chains engage in multivalent hydrophobic and π–π interactions (liquid–liquid phase separation / cohesive hydrogel formation, particularly in cohesive GLFG repeats of Nup98):

1. Entropic and Cohesive Exclusion of Inert Macromolecules: For an ordinary hydrophilic protein > 40 kDa lacking FG-binding pockets to enter the pore, it must break hundreds of cohesive Phe–Phe hydrophobic contacts inside the FG meshwork without forming compensatory bonds—imposing a prohibitive free-energy barrier (ΔG_entry ≫ k_BT).
2. Selective Dissolution ("Selective Phase") of Karyopherin-β (HEAT-Repeat) Transport Receptors: By contrast, every member of the Importin-β / Exportin (Kap-β) superfamily is folded as a flexible spring of 19 to 21 tandem HEAT α-helical repeats whose convex outer surface is studded with 8 to 10 shallow hydrophobic grooves. As an Importin-β–cargo complex approaches the pore, its surface hydrophobic pockets bind transiently (K_d ≈ millimolar per single FG motif, with sub-microsecond on/off kinetics) to the Phe rings of the FG-nucleoporins, locally "melting" the Phe–Phe cross-links and allowing the entire cargo complex to dissolve into and diffuse rapidly through the FG phase!

Because diffusion through the FG-nucleoporin hydrogel is thermally driven and thermally reversible, directionality and accumulation against a concentration gradient are powered entirely by the spatial compartmentation of the Ran GTPase cycle:

• Inside the Nucleus (High [Ran-GTP]): The guanine nucleotide exchange factor for Ran—RCC1 (Regulator of Chromosome Condensation 1, a 7-bladed β-propeller)—is permanently tethered to chromatin nucleosomes via direct binding to the acidic patch of histones H2A and H2B. Consequently, RCC1 continuously converts nuclear Ran-GDP into Ran-GTP inside the nucleoplasm.
• In the Cytoplasm (High [Ran-GDP], [Ran-GTP] ≈ 0): Conversely, RanGAP1 (Ran GTPase-Activating Protein 1) and its co-activators RanBP1 / RanBP2 (Nup358) are restricted strictly to the cytoplasm and cytoplasmic filaments of the NPC, where they accelerate Ran's intrinsic GTPase hydrolysis rate by 10⁵-fold, immediately converting any Ran-GTP that exits the nucleus into Ran-GDP + P_i (which is then recycled back into the nucleus by the dedicated homodimeric nuclear import factor NTF2 [Nuclear Transport Factor 2]).

This steep ~200:1 to 1,000:1 [Ran-GTP] gradient drives opposite allosteric switches in Importins versus Exportins:

1. Nuclear Import Cycle (Importin-α/β1 — Disassembled by Nuclear Ran-GTP):
   • In the cytoplasm (where [Ran-GTP] is virtually zero), Importin-α binds the cargo's classical monopartite (K-K/R-X-K/R, as in SV40 Large T-antigen PKKKRKV) or bipartite (K/R-K/R-X₁₀₋₁₂-K/R₃₋₅, as in nucleoplasmin and nuclear hormone receptors) Nuclear Localization Signal (NLS) inside its concave Armadillo groove, while Importin-β1 binds Importin-α's N-terminal IBB (Importin-β-Binding) domain.
   • The trimeric Cargo–Importin-α–Importin-β1 complex glides through the NPC FG-hydrogel into the nucleus.
   • Upon emerging into the nucleoplasm, abundant Ran-GTP binds with nanomolar affinity inside the N-terminal HEAT arch (HEAT 1–8) of Importin-β1, triggering a conformational change that ejects the IBB domain of Importin-α; simultaneously, the nuclear basket protein Nup50 catalyzes release of the NLS cargo into the nucleoplasm, while Ran-GTP–Importin-β1 and CAS (CSE1L)–Ran-GTP–Importin-α return to the cytoplasm where RanGAP1 hydrolyzes GTP to reset the cycle.
2. Nuclear Export Cycle (Exportin-1 / CRM1 — Assembled Cooperatively by Nuclear Ran-GTP):
   • In the nucleoplasm, unliganded Exportin-1 (CRM1) has very low affinity for a cargo's leucine-rich Nuclear Export Signal (NES, consensus Φ₁-X₂₋₃-Φ₂-X₂₋₃-Φ₃-X-Φ₄) because CRM1's hydrophobic NES-binding cleft is closed in the apo state.
   • Binding of nuclear Ran-GTP to the interior of the CRM1 HEAT ring allosterically opens the NES-binding cleft on the outer surface (increasing NES affinity by > 500-fold), forming a stable cooperative Cargo–CRM1–Ran-GTP ternary export complex.
   • After translocating through the NPC to the cytoplasmic filaments (Nup214/Nup358), RanGAP1 + RanBP2 stimulate Ran to hydrolyze GTP → GDP + P_i, causing the CRM1 ring to snap back to its low-affinity state and release the exported cargo into the cytosol.

How does the majority of bulk cellular mRNA exit the nucleus without using Ran-GTP or Exportins?

Whereas small non-coding RNAs (tRNAs via Exportin-t, pre-miRNAs via Exportin-5, and rRNAs via CRM1/NMD3) use canonical Ran-GTP-dependent exportins, the vast majority of bulk spliced messenger RNAs (mRNAs) exit the nucleus via an ATP-driven, Ran-independent pathway mediated by the heterodimeric export receptor NXF1–NXT1 (yeast Mex67–Mtr2). During transcription and RNA splicing, the Transcription-Export (TREX) complex and the Exon Junction Complex (EJC) deposit NXF1–NXT1 dimers along the mature mRNP. NXF1–NXT1 contains FG-binding domains that shuttle the mRNP through the NPC FG meshwork. When the mRNP reaches the cytoplasmic filaments (Nup214/Gle1), the ATP-dependent DEAD-box RNA helicase DDX19 (Dbp5)—activated specifically on the cytoplasmic face by Gle1 + inositol hexakisphosphate (IP₆)—hydrolyzes ATP to remodel the mRNP and strip off NXF1–NXT1, acting as a molecular Brownian ratchet that prevents the mRNA from sliding backward into the nucleus.

Why is Exportin-1 (CRM1 / XPO1) a major clinical target in cancer therapy?

Many hematologic malignancies (such as multiple myeloma and diffuse large B-cell lymphoma) and solid tumors overexpress Exportin-1 (XPO1 / CRM1) by 2- to 5-fold to continuously pump master tumor suppressor proteins—including p53 (TP53), Retinoblastoma (Rb), FOXO, p21, p27, BRCA1, and IκBα—out of the nucleus into the cytoplasm where they cannot activate cell-cycle arrest or apoptosis. Whereas the natural bacterial polyketide Leptomycin B covalently alkylates Cys528 inside the hydrophobic NES-binding groove of XPO1 irreversibly (causing severe clinical systemic toxicity), structure-based drug design yielded Selective Inhibitors of Nuclear Export (SINE) such as Selinexor (KPT-330, FDA-approved for relapsed/refractory multiple myeloma and DLBCL), which forms a slowly reversible covalent Michael-addition bond with Cys528, trapping tumor suppressors inside the cancer cell nucleus to trigger selective cancer cell apoptosis while sparing normal cells.

1. Görlich, D., & Kutay, U. (1999). "Transport between the cell nucleus and the cytoplasm." Annual Review of Cell and Developmental Biology, 15(1), 607–660. https://doi.org/10.1146/annurev.cellbio.15.1.607
2. Devos, D., Dokudovskaya, S., Alber, F., Williams, R., Chait, B. T., Sali, A., & Rout, M. P. (2004). "Components of coated vesicles and nuclear pore complexes share a common molecular architecture." PLoS Biology, 2(12), e380. https://doi.org/10.1371/journal.pbio.0020380
3. Frey, S., Richter, R. P., & Görlich, D. (2006). "FG-rich repeats of nuclear pore proteins form a three-dimensional meshwork with hydrogel-like properties." Science, 314(5800), 815–817. https://doi.org/10.1126/science.1132516
4. Mosalaganti, S., Obarska-Kosinska, A., Siggel, M., Taniguchi, R., Turoňová, B., Zimmerli, C. E., ... & Beck, M. (2022). "AI-based structure prediction empowers integrative structural analysis of human nuclear pores." Science, 376(6598), eabm9506. https://doi.org/10.1126/science.abm9506