The nuclear pore complex is the organelle within the nuclear envelope that mediates the export of messenger RNA, transfer RNA, and the subunits of ribosomes, which are synthesized or assembled in the nucleus but function in the cytoplasm. The pore complex also mediates the nuclear import of proteins and protein complexes (eg, snRNP particles) that are synthesized and assembled in the cytoplasm but destined to function in the nucleus. The number of nuclear pore complexes varies considerably, depending on the cell type. Their number generally reflects the biosynthetic activity of the cells. For example, the nuclear envelope of the Xenopus oocyte has the highest known concentration of pores per unit area of nuclear envelope (60 pores mm ), whereas quiescent chick erythrocytes have two to four pores mm . 1. Pore Structure
Nuclear pore complexes were first described by Callan and Tomlin (1) in their electron microscopic analysis of the nuclear envelope in amphibian oocytes. Architecturally, the nuclear pore is a huge (124 megadaltons), beautifully elaborate, symmetrical organelle within the nuclear envelope (Fig. 1). The entire complex can be graphically divided into four main component types: (1) eight-membered ring complexes, ( 2) the central plug, (3) cytoplasmic filaments, and (4) a filamentous nuclear basket (cage). There are two similar eight-membered ring complexes per pore, which associate through apparently homotropic interactions to create twofold symmetry of the complex within the plane of the envelope and are designated as the nucleoplasmic and cytoplasmic rings. The ring complexes are anchored to the nuclear envelope where the inner and outer membranes fuse together. Each ring consists of eight spoke-ring subunits of 10 to 25 nm arranged in eightfold symmetry when viewed from the surface of the envelope (3). Each spoke-ring segment consists of subcomponents that contribute to the formation of the overall complex. Some contribute to formation of the rings, and columnar and lumenal portions of each spoke-ring segment help anchor the ring complexes within the nuclear envelope. A spoke protrudes from each columnar subcomponent to help form the central annulus. Spokes from the cytoplasmic ring interact with spokes from the nucleoplasmic ring. The central annulus forms an hourglass-shaped channel of 42 nm at the peripheries and between 9 and 10 nm at the narrowest point. This channel transports macromolecular complexes as large as 23 nm (see later). Eight aqueous channels form on the periphery of the complex between the columnar subcomponents and the nuclear envelope. These aqueous channels permit passive diffusion of ions and small molecules across the envelope. Extending into the cytoplasm from the cytoplasmic ring are eight long, cylindrical filaments that may serve as docking sites for proteins that must enter the nucleus. Extending into the nucleoplasm from the nucleoplasmic ring are eight filaments that join together at their ends by yet another component to form a distal annulus within the nucleoplasm. This annulus may contact the nuclear matrix to play a role in the export of nuclear products.
Figure 1. A schematic model of the vertebrate nuclear pore complex. The cytoplasmic face is shown uppermost, and certain facing portions have been omitted for clarity. Structures are drawn approximately to scale with the pore transversing the membrane of ~90 nm thickness. The diameter of the outer spoke ring (OS) is ~120 nm. Filamentous structures (CF and NB) extend from both the cytoplasmic and nuclear faces, with the former attached to particles (P) and the latter forming a basketlike structure. The overall architecture is octagonal: based on eight equally spaced spokes (S) that are sandwiched between two rings (CR and NR). The spokes are attached to an inner spoke ring (IS) that encompasses a central plug (CP). The buttresses connecting the CP with the IS are omitted. The spokes appear to transverse the pore membrane for connection to an outer ring (OS) in the nuclear envelope (NE) lumen. Well defined lamina (L) and nuclear envelope lattice (NEL) structures also seem to have connections to the nuclear pore complex.
At least 100 different proteins (nucleoporins) are thought to constitute the nuclear pore complex, and approximately forty have been identified. They can be subdivided into pore membrane proteins (poms) and phenylalanine/glycine-rich (FG) proteins. Poms are integral nuclear membrane proteins closely associated with the pore complex, such as the 16 to 24 copies of protein gp210 that reside around the periphery of the pore complex. They anchor the pore complex within the envelope, and the bulk of the gp210 resides within the lumen of the envelope. Other poms include the yeast Pom152p and the mammalian pom121. FG proteins found in yeast, such as NUP1, NUP2, and NSP1, are so-called because of their repeated -X-Phe-X-Phe-Gly- (or -XFXFG-) sequences. A related -Gly-Leu-Phe-Gly- (or -GLFG-) repeated sequence is also found in yeast proteins NUP100, NUP116, NUP49, and NUP145 . The functions of the repeated motifs remain unknown. Deleting them does not perturb the cell’s viability nor localization of the protein to the complex. Several metazoan nucleoporins also have been identified. Some of these have been localized to the spoke-rings that face the cytoplasm and their extending filaments (nup214/CAN, p180, and p75), some to the central plug or spoke-ring structures (p62, p58, p54), and at least one to the nuclear basket (nup153). Interestingly, outside of the FG-repeat domains, no extensive homologies have been found between yeast and metazoan nucleoporins. Metazoan FG nucleoporins are modified posttranslationally by addition of N-acetylglucosamine through O-glycosylation of serine and threonine residues. The function of this glycosylation also remains unknown, but it makes it possible to probe specifically for nucleoporins and to block transport using the wheat germ agglutinin (WGA), a lectin that specifically binds O-linked N-acetylglucosamine.
2. Pore Function
The pore complex transports macromolecules into and out of the nucleus and permits diffusion of small molecules across the nuclear envelope (see Nuclear Import, Export). Briefly, proteins that are destined for the nucleus translocate across the envelope by a facilitated diffusion mechanism, as in the case of small proteins such as histones (4), or an active transport, ATP-driven mechanism, as in the case of proteins larger than approximately 40 kDa. Large proteins usually have a short nuclear localization signal (NLS) within their linear amino acid sequence. Two types of signals are known. One is a simple basic sequence, rich in lysine residues, typified by the NLS found in the SV40 virus large T Antigen(namely, -Pro-Pro-Lys-Lys-Lys-Arg-Lys-Val-). Fusion proteins that consist of a normal cytoplasmic protein linked to this NLS translocate into the nucleus. Another characterized NLS is a bipartite sequence in which two short segments of basic residues are separated by approximately ten amino acid residues (5).
Transport of nuclear proteins harboring NLS sequences occurs in two steps (6). The first step is an energy-independent docking that occurs initially with the cytoplasmic filaments extending from the envelope pore complex and then with the central annulus. Cytoplasmic factors, known as importins a and b (7, 8), bind NLS sequences of proteins destined to function in the nucleus and deliver them to the filaments. Association occurs between the importins and the FG nucleoporins described previously. The docking protein may be nup214 of the filaments. Docking can be blocked by WGA, which specifically binds the O-linked #-acetylglucosamine residues on nucleoporins that constitute either the filaments (ie, nup214) or the central annulus of the pore complex. The second step is translocation of the docked proteins, which requires GTP hydrolysis (9-11). Translocation requires Ran/TC4, a Ras-like GTPase, and its specific guanine nucleotide exchange factor, called RCC1. The precise mechanism of the active transport remains uncertain, however, and other factors may be required, such as the recently described nuclear factor II (NTFII), which transiently associates with the pore complex protein p62 and Ran/TC4 (12). ATP hydrolysis is necessary for nuclear protein uptake, but again, the actual reason remains uncertain.
The nuclear pore complex is also required for nuclear export of messenger RNA (mRNA), transfer RNA (tRNA), ribosomal subunits, and prespliceosomal complexes that mature in the cytoplasm before reentering the nucleus (13, 14). The export mechanisms of these gene products are still largely unknown. Although export of mRNA must include docking, unwinding, and then translocation, it is known that the mRNA leaves the nucleus bound to protein. The nuclear cage must be intimately involved in the export process (15). Blobel’s "gene gating" hypothesis (16) suggests that nuclear pores and the nuclear lamina influence (by specific binding of chromatin) the three-dimensional gene organization within the nucleus, so that transcribed genes are positioned for efficient transport of their RNA products through the pores. Tracks of polyA-containing mRNA have been detected in nuclei between the sites of RNA synthesis and the nuclear pores, lending support to the gating hypothesis. These tracks may be the perichromatin fibrils, which contain several RNA splicing components (17). Despite these uncertainties, pre-mRNA does not interact with the pore complex until all the splicing reactions are complete. HnRNP proteins (eg, A1, A2, and B1) may play a positive role in this export because a non-RNA-binding domain has been identified in their linear amino acid sequence as a nuclear export signal (NES) (14). When fused to unrelated proteins, the NES in A1 (called M9) confer bidirectional import and export (shuttling) of the fusion protein (18, 19). Other nuclear proteins contain export signals that are unrelated to the M9 sequence. For example, the activation domain of the HIV Rev protein is leucine-rich and is a strong NES that mediates the export of Rev with its bound, unspliced HIV RNA (14). The NES in Rev is similar to the NES in transcription factor IIIA (TFIIIA), which binds and exports 5S rRNA (20).
The nuclear pore complex disassembles into individual components during mitosis. Pore disassembly is concomitant with mitotic disassembly of the nuclear lamina and the vesiculation of the envelope membranes. Pore reassembly begins in telophase and is initiated by nucleoporins and poms aggregating within the plane of the re-forming envelope. Pore assembly itself is probably incremental, and individual components arrive at the site of reassembly from a stored mitotic pool. Fusion of the inner and outer membranes occurs at the site of pore assembly. Later in the cell cycle during the S-phase, pore complexes are added as the nuclear envelope grows in size. Again addition may be incremental as in telophase. Alternatively, cytoplasmic annulate lamellae, which consist of membrane and pore-like structures, may contribute additional membrane and pores to the growing nuclear envelope.
