sign in sign up
<<
Home , Nuclear pore

Journal of Cell Science 113, 1651-1659 (2000) 1651 Printed in Great Britain © The Company of Biologists Limited 2000 JCS0712

COMMENTARY The nuclear pore complex: mediator of translocation between nucleus and cytoplasm

T. D. Allen1, J. M. Cronshaw1, S. Bagley1, E. Kiseleva2 and M. W. Goldberg1 1CRC Structural Cell Biology Group, Paterson Institute, Christie Hospital, Manchester, M20 4BX, UK 2Institute of Cytology and Genetics, Russian Academy of Science, Novosibirsk, 630090, Russia

Published on WWW 18 April 2000

SUMMARY

The enclosure of nuclear contents in means that receptors and adapters, and the molecules (and their cells require sites in the boundary that mediate exchange regulators) that underpin the transport mechanisms. Over of material between nucleus and cytoplasm. These sites, the past few years there has been an increasing interest in termed nuclear pore complexes (NPCs), number 100-200 in the pore complex: structural studies have been followed by yeast, a few thousand in mammalian cells and ~50 million elucidation of the biochemical aspects of nuclear import, in the giant nuclei of amphibian oocytes. NPCs are large and subsequent investigations into nuclear export. The (125 MDa) macromolecular complexes that comprise current challenge is to understand the interactions between 50-100 different in vertebrates. In spite of their the structural elements of the pore complex and the size and complex structure, NPCs undergo complete mechanisms that drive the physical processes of breakdown and reformation at cell division. Transport translocation through it. through NPCs can be rapid (estimated at several hundred molecules/pore/second) and accommodates both passive of relatively small molecules, and Movies available on-line: of complexes up to several megadaltons in molecular mass. http://www.biologists.com/JCS/movies/jcs0712.html Each pore can facilitate both import and export. The two processes apparently involve multiple pathways for Key words: Nuclear pore complex (NPC), Nucleocytoplasmic different cargoes, and their transport signals, transport transport, NPC structure, NPC function, NPC dynamics

INTRODUCTION , including the sequencing of all 30 nucleoporins found in isolated yeast NPCs has been published recently (Rout The vertebrate nuclear pore complex (NPC) is one of the et al., 2000). Fontoura et al. (1999) have initiated similar largest types of macromolecular complex in the cell: it has an approaches for the 50 nucleoporins isolated from rat liver estimated molecular mass of 125 MDa in vertebrates (Reichelt NPCs. The next major task is a full structural characterisation et al., 1990), and comprises 50-100 proteins (Fontoura et al., of the nucleoporins, their exact localisation within the overall 1999) termed nucleoporins. Yeast pore complexes are smaller, NPC structure and definition of the specific roles of these being made up of ~30 nucleoporins and having a molecular proteins in transport (Fig. 1). mass of ~66 MDa (Rout and Blobel, 1993; Rout et al., 2000). Amongst the variety of nucleoporins that make up the pores, several complexes have been identified, such as the p62 STRUCTURE OF THE NPC complex in vertebrates, which consists of p62, p58 and p54 in Xenopus (Finlay, 1991) as well as p45 in the p62 complexes Yeast and vertebrate NPCs display an overall similarity of core from rat (Guan et al., 1995). Several complexes structural elements at the electron microscopic level (Yang et have been described in yeast (see Ohno et al., 1998); a recently al., 1998), and equivalent yeast and vertebrate nucleoporins described example, the Nup84 complex (Rappsilber et al., share sequence similarity of ~25%. The increased size of the 2000), has also been structurally characterised as a triskelion vertebrate pore is most likely to reflect an increase in that might form an elemental building block of the pore complexity consistent with metazoan evolution, but the basic structure. Excellent accounts dealing with the nucleoporins structural similarities between the NPCs of yeast and higher themselves can be found in other recent reviews (Ohno et al., eukaryotes indicate that they probably share common 1998; Stoffler et al., 1999a). A full characterisation of the yeast translocation mechanisms. Extensive localisation studies of 1652 T. D. Allen and others yeast NPCs by immunoelectron microscopy have shown a peripheral structures of the NPC. Transport (both import and symmetrical distribution of potential transport-associated export) is then completed by a series of cargoÐcomplex-NPC docking sites running in a continuous sequence through the interactions over a distance of up to 200 nm, required for pore structure (see Stoffler et al., 1999a; Rout et al., 2000). complete passage between nucleus and cytoplasm (Ohno et al., The overall structure of the vertebrate nuclear pore has been 1998). known for some time (Akey, 1989; Hinshaw et al., 1992; Goldberg and Allen, 1993, 1996; Panté and Aebi, 1996; Goldberg et al., 1999; Kiseleva et al., 1998). The presence of , SOLUBLE FACTORS AND a central plug or transporter (Akey and Radermacher, 1993; Goldberg and Allen, 1996) as a fundamental element of pore structure remains controversial: an alternative view suggests Molecules destined for nuclear import (Fig. 3) or export (Fig. that central elements are merely cargo caught ‘in transit’ during 4) can be thought of as cargoes, and include proteins, various specimen preparation. Our own view is that a regular and types of RNA and RNPs. For recent comprehensive reviews, consistent structure is visualised in all well-prepared NPCs, see Nakielny and Dreyfuss (1999) and Görlich and Kutay either rapidly frozen (Akey, 1989) or chemically fixed, and can (1999). Crucial to the process of transport are the import be exposed by proteolysis (Goldberg and Allen, 1993, 1996). receptors (Görlich et al., 1995) such as α and importin This view is also supported by the direct visualisation of the β, the earliest-identified members of a family comprising at transporter as a central NPC structure through which mRNP least seven members (Mattaj and Engelmeier, 1998). In yeast fibres emerge into the cytoplasm during mRNA export in the import receptors are termed . Export receptors Chironomus salivary glands (Kiseleva et al., 1998). Studies of (exportins) recognise nuclear-export signals (NESs) as specific the passage of gold particles through the centre of the pore sequences in the same way as nuclear-localisation signals complex by transmission EM indicate a central channel (NLSs) are recognised by . Further molecules (through the transporter) that has a maximum diameter of 26 necessary for transport include the small GTPase Ran and nm (Feldherr and Akin, 1993, 1994a,b). In the absence of any NTF2 (nuclear-transport factor 2). central structure in the pore complex, a considerably larger In a typical example of import, the NLS of nucleoplasmin central orifice (perhaps up to 45 nm) would be expected. is recognised by importin α. Importin α becomes complexed Our current view is that the transporter is suspended in the to importin β, and the cargo plus importin α and importin β is centre of the spoke-ring complex, which occupies most of the translocated through the NPC into the nucleus, where it is space delineated by the pore membrane (the area of membrane that joins the inner and outer nuclear envelopes). The radial arms of the spoke complex penetrate the pore membrane and project into the lumen of the . Sandwiching the spoke- ring complex are the cytoplasmic coaxial ring and the nucleoplasmic coaxial ring, which anchor and support the cytoplasmic filaments and nuclear baskets, respectively (Fig. 2A,B). The cytoplasmic coaxial ring has at least three levels of substructure. It contains a ‘star ring’ in direct apposition to the outer NE membrane; the star ring itself is covered by a thin ring, to which further elements are added to form the entire ring, and this supports the attachment of eight cytoplasmic filaments (Fig. 2B). The cytoplasmic coaxial ring is joined to the entrance of the transporter by eight internal filaments, which are arranged rather like the hub and spokes of a wagon wheel. These substructures were originally identified by FEISEM after various fracturing and proteolytic treatments (Goldberg et al., 1996) and subsequently confirmed in successive stages of NPC assembly in vitro (Goldberg et al., 1997). Although 95% of the NPC’s in isolated oocyte envelopes show typical eightfold radial symmetry, a small proportion of examples of nine- and seven-fold symmetry are seen, often adjacent to one another; this might indicate local disruption of a self-assembly process. This is particularly evident in the basket structures, in which the increase in diameter required to accommodate nine rather than eight structural elements is clearly apparent (Goldberg and Allen, 1993). Nuclear transport is initiated by interaction of transport cargo with the Fig. 1. Functions of the nuclear pore complex. The nuclear pore complex 1653 dissociated from the importins by GTP-bound Ran. Ran-GTP rapid isolation and fixation of oocyte nuclear envelopes binds to importin β and recycles it back to the cytoplasm, revealed a much longer filamentous organisation (Goldberg where GTP hydrolysis of Ran is stimulated by the GTPase and Allen, 1996); this is also apparent in situ in transmission activating Ran-GAP; this releases importin β for the next round electron microscopy (TEM) when gold-labelled transport of import. The new Ran-GDP is taken back into the nucleus complexes are bound to the filaments (Richardson et al., 1988; by NTF2 (Fig. 5), where RCC1 stimulates exchange of Ran- Rutherford et al., 1997). There have also been suggestions that bound GDP for GTP, which effectively converts Ran-GDP elements of the cytoskeleton make direct contact with these back to Ran-GTP. Meanwhile importin α and the cargo have separated in the nucleus, and importin α is recycled to the A cytoplasm complexed to CAS (the product of the cellular apoptosis-susceptibility gene) and Ran-GTP (Kutay et al., 1997). Thus Ran-GDP is thought to exist at a relatively higher concentration in the cytoplasm, whereas a high concentration of Ran-GTP exists in the nucleus. Nuclear import is clearly linked to this Ran-GTPase cycle (Fig. 5), although it is not necessarily dependent on the actual process of nucleotide hydrolysis. Two fundamental aspects of transport remain unclear. Firstly, the requirement for cargoes to perceive which side of the nuclear envelope is nuclear and which is cytoplasmic. The current view is that there are high concentrations of Ran- GDP in the cytoplasm, and high concentrations of Ran-GTP in the nucleus. High concentrations of Ran- B GDP promote the formation of import complexes, whereas high concentrations of Ran-GTP dissociate them. Conversely, high concentrations of Ran-GTP promote the formation of export complexes, and high concentration of Ran-GDP cause their dissociation. Ran cycles through the pores with both complexes, being converted to Ran-GTP by RCC1 in the nucleus, and to Ran-GDP by Ran-BP and Ran-GAP, both of which localise on the cytoplasmic filaments of the nuclear pore complex. The import and export pathways and the Ran cycle from this commentary are available as animations at http://www.biologists.com/JCS/. Secondly, although the actual mechanisms of translocation through the pore are theoretical at best, some of the structures and their potential roles are considered below.

NUCLEOPORINS IMPLICATED IN TRANSPORT AND THEIR LOCALISATION

Cytoplasmic NPC structures Fig. 2. (A) Schematic representation of the nuclear pore complex. A cutaway representation of Cytoplasmic filaments visualised by the nuclear pore complex as seen from an axis perpendicular to the NE. It should be noted that FEISEM are usually short stubby the NPC has 8-fold rotational symmetry. (B) Structural respresentation of the NPC with direct structures (Goldberg and Allen, 1993), but visualisation of individual components by FEISEM (modified after Goldberg et al., 1999). 1654 T. D. Allen and others

filaments and provide a potential route from deep in the increase in the diameter of the channel in transformed cytoplasm to the nuclear surface (Goldberg and Allen, 1995). mammalian tissue culture cells (Feldherr and Akin, 1995). Whether import cargoes are delivered via cytoskeletal elements The mechanism of entry to (and exit from) the channel or not, the nuclear pore cytoplasmic filaments provide the through the transporter remains uncharacterised, and a recent initial docking site for active nuclear import. finding by Feldherr and Akin (1997) suggests that the channel The evidence to date indicates that Nup358, and has a single ‘gate’ in the centre. When transport is blocked with CAN/Nup214 are two of the major cytoplasmic filament WGA gold, however, no nucleoplasmin gold is apparent within nucleoporins (Fig. 6). Nup358 (also termed RanBP2) is the channel; this suggests exclusion of the gold at the surface currently the largest vertebrate nucleoporin identified (Wu et of the NPC, and implies that there is a gate in this region al., 1995; Yokoyama et al., 1995). Nup358 has eight zinc- (Rutherford et al., 1997). WGA labelled with gold binds over finger motifs, which serve as novel Ran-binding domains, the surface of the internal filaments, which radiate from the binding exclusively to Ran-GDP. It also has four other Ran- centre of the pore at the level of the nuclear envelope. These binding sites, which are distinct from the zinc fingers. These internal filaments are a constant feature of pore complexes Ran BP1 homology (RBH) sites bind both Ran-GTP and Ran- imaged by FEISEM (Goldberg and Allen, 1996, 1999) and GDP, but have a much higher affinity for Ran-GTP. EM might open or close the channel at the entrance to the studies using Ran complexed to gold colloid indicate that the zinc-finger domains are in the more peripheral parts of the cytoplasmic filaments (Yaseen and Blobel, 1999). Nup358 is likely to be involved in transport, because antibodies to it block import (Yokayama et al., 1995), and the characteristic nucleoporin FXFG repeats (where X is an amino acid with a small or polar side chain) bind importin β (Delphin et al., 1997). Nup214 (also called CAN) is located at the base of the cytoplasmic filaments, and might also contribute also to the cytoplasmic coaxial ring (Panté et al., 1994; Bastos et al., 1997). Nup214 forms complexes with Nup88, Nup84 and probably p62 (Macaulay and Forbes, 1996). Nup214-deficient mice have impaired nuclear transport (Fornerod et al., 1997). Despite its cytoplasmic location, Nup214 might play a role in export, given that it co-immunoprecipitates with exportin 1. Because of its cytoplasmic localisation, Nup214 probably functions in a terminal step (Fornerod et al., 1996). The yeast homologue Nup159p (Gorsch et al., 1995; Kraemer, 1995) forms a complex with Nup82p that is also found on the cytoplasmic face (Grandi et al., 1995), and mutations in either gene result in inhibition of poly(A) RNA export but not import (Belgareh et al., 1998). Thus, peripheral nucleoporins can affect import or export mechanisms initiated at the ‘opposite’ side of the NPC. These somewhat distant interactions may reflect a mechanism of vectorial transport in which cargo complexes have increased affinities for a distal region of the pore complex. Structures in the centre of the pore Running through the centre of the transporter, and providing the route of all actively translocating cargoes, is a central channel that can vary in diameter from potentially closed to 25 nm (measured by translocation of colloidal gold coated with nucleophilic ; Feldherr and Akin, 1993, 1997). This maximum diameter is also required for the export of Balbiani ring Fig. 3. Nuclear import. A schematic representation of nuclear import through (BR) RNP complexes from polytene chromosomes in the NPC. The transport receptor and import substrate form a complex in the cytoplasm where RanGTP levels are low (Importin a and b are not shown Chironomus salivary glands (Kiseleva et al., 1998). In separately for simplicity). This complex is translocated through the NPC. contrast, the maximum channel diameter in amoebae is Nuclear RanGTP induces the dissociation of the import cargo from its only 13 nm (Feldherr and Akin, 1999). The diameter receptor (and, in some cases, dissociation of the receptor from the NPC). The of the channel also varies following oncogenic transport receptor is recycled to the nucleus, probably as a complex with transformation, which generates a small but significant RanGTP. The nuclear pore complex 1655 transporter (Kiseleva et al., 1998). A second set of these (5) If the NPC has gating properties, are they directly involved filaments is found on the nucleoplasmic side of the transporter in ‘steady state’ transport? (Goldberg et al., 1999); these might function similarly as a gating mechanism. Between the two sets of filaments, and Structural alterations in the NPC surrounding the transporter, lies the spoke ring complex (Akey Several variations in NPC structure have been reported, and Radermacher, 1993) consisting of the inner spoke ring, generated from a variety of technical approaches. Akey showed which is surrounded by a transmembrane domain to which that rapidly frozen unfixed NPCs exist in different forms, Pom121 has been localised (Hallberg et al., 1993). One of the namely ‘open’, ‘in transit’ or ‘closed’ configurations (Akey, earliest characterised nucleoporins, gp210, localises to the ends 1990). The functional size of the central NPC channel increases of the eight radial arms that protrude into the lumen of the at later stages of Xenopus oocyte development and can be nuclear envelope (Greber et al., 1990). Despite this apparent modulated by microinjection of NTF2 (nuclear transport factor; separation from the central channel, antibodies to gp210 inhibit Feldherr et al., 1998). The position of the transporter relative nuclear transport (Greber and Gerace, 1992). to other NPC components varies during mRNP export in The internal filaments on the nucleoplasmic side appear to Chironomus (Kiseleva et al., 1998), as shown by FEISEM. lie within the nucleoplasmic star ring, where Nup155 has been Independent approaches by atomic force microscopy and located (Radu et al., 1993) along with the p62 complex and also Nup93 (Grandi et al., 1995). The role of these structures (if any) in transport is currently unknown, but their localisation in the pore makes it a reasonable possibility. Nup 153 is unusual in that it contains at least four zinc-finger motifs, which are usually involved in binding to DNA or RNA. For Nup153, however, it is more likely that they are involved in protein-protein interactions most probably in the binding of Ran in its GDP bound form (Nakielny and Dreyfuss, 1999). The same authors also provided interesting evidence that Nup153 shuttles between the nuclear and cytoplasmic faces of the pore complex. Nup153 also associates with at an early stage of nuclear envelope reassembly at (Bodoor et al., 1999), and might be involved in chromatin organisation as an intranuclear filament. Several of the currently characterised vertebrate nucleoporins have been localised to the nucleoplasmic basket (Fig. 6). The basket provides the last NPC structure encountered by imported cargoes and the first encountered by exported cargoes, although there is evidence for attachment of filamentous proteins to the basket, and these might extend deep into the nuclear interior both in vertebrates (Tpr, Nup98, Nup153, see Stoffler et al., 1999a) and in yeast (Kosova et al., 1999). Thus, the NPC might provide a physical link between elements of the cytoskeleton in the cytoplasm, and fibrous proteins that could run throughout the entire nucleus and potentially interconnect the whole of the cytoplasm with the whole of the nucleus. We can now address the following questions with respect to NPC structure and function in the transport process. (1) Is the NPC rigid or capable of structural alteration, and are conformational changes required for the transport mechanisms? (2) How does the NPC contribute to the transport process? Is there a series of binding sites between which cargo complexes move, or is there a structural alteration of the pore during transport that contributes to the Fig. 4. Nuclear export. A schematic representation of nuclear export through mechanism of transport? the NPC. The transport receptor and export substrate form a complex in the (3) Do ‘large’ cargoes generate pore complex responses nucleus with RanGTP. This complex is exported from the nucleus and different from those generated by smaller cargoes? dissociated in the cytoplasm by hydrolysis of Ran-bound GTP to GDP. This (4) Does diffusion proceed undiminished regardless of releases the export substrate and the transport receptor is recycled into the any other activities in process in an individual pore? nucleus. 1656 T. D. Allen and others

FEISEM both showed variations in NPC basket structure same pore, possibly running parallel and facilitating transport induced by Ca2+ ions (Stoffler et al., 1999b) or during RNP in opposite directions. It has yet to be established whether export in Chironomus (Kiseleva et al., 1998). Although the active transport through NPCs proceeds in both directions at electron microscopic approach to observation of these changes the same time, whereas diffusion presumably does. naturally involved stabilisation of the NPC by fixation or rapid For a system in which transport through the NPC is achieved freezing, the changes observed were consistent with individual by a system of successive binding sites, cargo complexes samples. Furthermore, the atomic force microscopy was carried would bind at the most peripheral parts of the NPC, and then out on unfixed, hydrated specimens, and the effects of Ca2+ separate from this original site to bind again, successively, were reversible (Stoffler et al., 1999b). deeper and deeper into the NPC structure, finally emerging on Perhaps the most obvious structural alterations of the NPC the other side. The individual components would uncouple at associated with transport are those involved in export of the the last stage in this process, allowing recycling of the BR granule (of mRNP) in Chironomus. Initially, the NPC transporter molecules. If a serial binding mechanism exists, basket acts as a docking site, orientating the 50-nm BR granule several questions arise: how many binding sites are there, how in such a way that the 5′ end of the mRNP leads the way far apart are they and where are they situated, and how is through the transporter as the BR granule essentially ‘unravels’ directionality maintained. to pass through the NPC as a 25-nm-diameter fibre. As this In a theoretical consideration of export, could the cargo bind unwinding of the BR granule proceeds, the diameter reduces, and the NPC basket accommodates to this change, continuing until the entire mRNP fibre passes through the pore; at this point, the basket organisation returns to a ‘resting’ configuration prior to the next round of mRNP export (Kiseleva et al., 1998). The mechanism by which the NPC basket ring can change its dimensions might be brought about by the basket fibres: each of these is formed from two twisted subfibres (Goldberg and Allen, 1996), which could unwind to allow an increase in basket ring diameter, or coil more tightly to reduce the basket ring diameter. Although this example of direct involvement of NPC peripheral structure might be considered to be unusual, given that it occurs in cells that have permanently condensed giant chromosomes, it clearly is normal physiology, and, as such, provides a useful paradigm for the role of the NPC basket in export of large mRNP cargoes, and possibly for other export mechanisms.

NPCs AS ‘SERIAL BINDING’ COMPLEXES

Recent work on biochemically separated yeast NPCs (Rout et al., 2000) has isolated 30 nucleoporins, and intensive EM immunocytochemistry has defined a mainly symmetrical distribution of nucleoporins about the plane of the nuclear envelope. This distribution of labelling gives an unbroken series of potential cargo-binding sites through the entire NPC. A current view is that a system of ‘binding- site tracks’, along which a transport cargo could move by repeated attachment (and detachment), might exist. Certain other parameters would need to be fulfilled for this type of mechanism to work, particularly in respect of establishing Fig. 5. The Ran cycle. The small GTPase Ran, in its GTP-bound form, is constantly directionality, which could be achieved by exported from the nucleus in complexes with transport receptors. RanGTP increased binding affinity in appropriate parts of hydrolysis in the cytoplasm is accomplished by RanGAP and RanBP1/RanBP2. the NPC. This type of mechanism could facilitate RanGDP can then bind its import receptor NTF2 which mediates its import into the bi-directional transport: different binding tracks nucleus. In the nucleus, RCC1 mediates exchange of GDP for GTP and restores Ran for import and export could be situated in the to its GTP-bound state. The nuclear pore complex 1657 initially at a nucleoporin at the NPC basket ring (Nup153), then formation in in vitro systems have revealed several stages transfer to a centrally positioned nucleoporin (the p62 during NPC formation: the first stage is a small (4-nm complex?) and then to a cytoplasmic filament binding site diameter) dimple in the outer nuclear envelope, interpreted to (Nup214/CAN?). At this stage, the export complex would be be the initial interaction between inner and outer nuclear dissociated, possibly at a more distal part of the cytoplasmic envelopes in the formation of the pore-membrane domain. This filaments (Nup358/RanBP2?), which would allow recycling of is followed by progressive incorporation of components that the transport factors. In the case of NFAT, a shuttling starts from the internal organisation of the pore and concludes factor, Ran-GTP is required for binding of with the formation of peripheral structures (Goldberg et al., CRM1 (the nuclear-export receptor) to the p62 complex and 1997; Gant et al., 1998). To date, the stages of formation have NuP214/CAN; the export complex is then dissociated by been observed (both in vivo and in vitro) only from the RanBP1 and RanBP2 (Nup358), which is situated at the tips cytoplasmic face by surface imaging, and stages in formation of the cytoplasmic filaments (Kehlenback et al., 1999). Would of nucleoplasmic pore structures have not been documented. 3-4 interactions suffice over a journey of 200 nm? During nuclear assembly in vitro, pores are formed as soon as Directionality could be maintained by subsequent occupation there is sufficient area of flattened nuclear envelope to of each binding site by the next cargo Ð given a suitable accommodate them Ð well in advance of complete nuclear concentration of cargo ‘queuing’ for transport. Alternatively, a envelope enclosure. Complete pore formation in vitro takes gradient of increasing numbers or affinities of binding sites place within six minutes. In vivo, however, in early Drosophila along the route could satisfy the requirements for embryo cell division, the rapidity of division cycles indicates directionality. even faster NPC assembly, despite the incomplete breakdown A further example of direct interaction between structural and reassembly of the nuclear envelope in this situation (E. elements of the NPC and transport-related molecules involves Kiseleva, personal communication). the nuclear import of Ran-GDP (Bayliss et al., 1999), which is Recent studies have indicated the order in which some dependent on NTF2, a homodimeric protein that binds to Ran- nucleoporins are incorporated during NPC formation. When GDP and also to nucleoporins that have XFXFG repeats, such the extracts used in in vitro nuclear assembly are depleted of as p62. Structural investigation shows that NTF2 binds through WGA-binding nucleoporins, pore formation is inhibited and the residues in the XFXFG sequences; it can be restored by readdition of the depleted proteins (Powers probably uses an exposed side chain of residue 7. The Ran- et al., 1997). When these proteins are separated and added back binding site of NTF2 is in a hydrophobic cavity in the centre individually, if the p62 complex is added alone, only early of the NTF2 molecule; Ran is situated rather like the ice-cream stages of pore structure result. Pore formation can be in a cone. The binding constants between NTF2 and the completed by the subsequent addition of the NuP214/CAN XFXFG repeats on the nucleoporins indicate rates of fraction. If the NuP214/CAN fraction is added back alone, no interaction of 100-1000 per second (Chaillan-Huntingdon et pore complexes form, which indicates the requirement for al., 2000). incorporation of p62 before Nup214/CAN (J. M. Cronshaw,

ROLE OF GATING

The extremely rapid rate of transport through the pore would appear to preclude a gating requirement as part of standard transport mechanisms. There may well be a role for NPC gating when requirements for nucleocytoplasmic transport are altered. These might include stages of both formation and disassembly of the nuclear envelope and pore complexes during cell division, as well as situations in which the requirement for transport is reduced (e.g. periods of quiescence, such as Go phase, or during pathological responses to trauma, such as heat shock, or early stages of apoptosis). Variations in NPC transport occur during the (Maul, 1977; Feldherr and Akin, 1994a), suppression of transport by p53 (Feldherr and Akin, 1994b) and in quiescent cells (Feldherr and Akin, 1993). Nuclear pore clustering is an early feature of apoptotic cells (Reipert et al., 1999), although whether nuclear transport is suppressed at this stage is unknown.

DYNAMICS OF NUCLEAR PORES

Despite its gross size and ornate molecular architecture, the NPC breaks down and reassembles at cell division in vertebrates within a fairly short period of time. Studies of NPC Fig. 6. Localisation of nucleoporins within the nuclear pore complex. 1658 T. D. Allen and others unpublished). Similarly, the order of recruitment of consequence has been the realisation that modulation of nucleoporins to the reforming nuclear envelope in early nuclear transport is a potential route to anti-viral therapy. anaphase in mammalian tissue culture cells involves NuP153 Given the two main systems outlined above, and the and POM121, followed at anaphase/ by the p62 increasing variety of import and export assay systems in higher complex, NuP214 and NuP84, and finally, at telophase, Gp210 eukaryotic systems initiated by Adam and Adam (1994), the and Tpr (Bodoor et al., 1999). prospects for further significant progress in our understanding Nuclear envelope breakdown and pore disassembly at cell of the mechanisms of nuclear transport would appear to be division is currently less well characterised, but there appears excellent. As well as eludication of a basic biological to be a requirement for almost complete removal of NPCs from phenomenon, the opportunity for control of nuclear import and the nuclear envelope before the membranes themselves begin export in clinical intervention are important areas for future to breakdown (Cotter et al., 1998). NPC disassembly in vitro studies. exhibits a series of stages that exactly match those of assembly (in reverse) Ð namely a sequential and progressive loss of T.D.A., S.B. and M.W.G. acknowledge support from the CRC (UK), peripheral components, followed by loss of the core E.K. the Wellcome Foundation, and J.M.C. the M.R.C. (UK). Many constituents of the pore and, finally, an overall re-sealing of the thanks to Elaine Mercer for preparing the manuscript. nuclear envelope membranes in vitro. At this stage, breakdown of the nuclear envelope itself proceeds either by tubule or REFERENCES vesicle formation in vitro (Cotter et al., 1998). Whether nuclear envelope breakdown in vivo involves resorption of membrane Adam, E. J. and Adam, S. A. (1994). Identification of cytosolic factors directly into the via tubules or involves required for nuclear location sequence-mediated binding to the nuclear vesicle formation by the nuclear envelope membranes remains envelope. J. Cell Biol. 125, 547-555. controversial. Akey, C. W. (1989). Interactions and structure of the nuclear pore complex revealed by cryo-electron microscopy. J. Cell Biol. 109, 955-970. Akey, C. W. (1990). Visualization of transport-related configurations of the nuclear pore transporter. Biophys. J. 58, 341-355. CONCLUSIONS AND PERSPECTIVES Akey, C. W. and Radermacher, M. (1993). Architecture of the Xenopus nuclear pore complex revealed by three-dimensional cryo-electron The past decade has seen a considerable increase in both microscopy. J. Cell Biol. 122, 1-19. Bastos, R., Ribas de Pouplana, R. L., Enarson, M., Bodoor, K. and Burke, interest in and investigation of the mechanisms of B. (1997). Nup84, a novel nucleoporin that is associated with CAN/Nup214 nucleocytoplasmic transport. A recent review covering yeast on the cytoplasmic face of the nuclear pore complex. J. Cell Biol. 137, 989- and vertebrates cites 16 categories of import signal and 1000. receptor, ten categories of export signal and receptor, and five Bayliss, R., Ribbeck, K., Akin, D., Kent, H. M., Feldherr, C., Gorlich, D. and Stewart, M. (1999). Interaction between NTF2 and xFxFGÐcontaining shuttling signals and receptors (Nakielny and Dreyfuss, 1999). nucleoporins is required to mediate nuclear import of Ran GDP. J. Mol. Biol. It is likely that these numbers will increase further. The power 29, 579-593. of yeast genetics has led to considerable progress in the field, Belgareh, N., Snay-Hodge, C., Pasteau, F., Dagher, S., Cole, C. N. and Doye although there are several yeast molecules involved in a nuclear V. (1998). Functional characterization of a Nup159p-containing nuclear pore pore structure and function that have no known vertebrate subcomplex. Mol. Biol. Cell 9, 3475-3492. Bodoor, K., Shaikh, S., Salina, D., Raharjo, W. H., Bastos, R., Lohka, M. homologues. The biochemical isolation procedure for yeast and Burke, B. (1999). Sequential recruitment of NPC proteins to the NPCs (Rout et al., 2000) has been followed by 2D-gel analysis nuclear periphery at the end of mitosis. J. Cell Sci. 112, 2253-2264. and sequencing of all 30 yeast nucleoporins in the isolated pore Chaillan-Huntingdon, C., Braslavsky, C. V., Kuhlmann, J. and Stewart, complex. Several yeast nucleoporins have been mutated and M. (2000). Dissecting the interactions between NTF2, RanGDP and the nucleoporin XFXFG repeats. J. Biol. Chem. 275, 5874-5879. expressed, often as temperature-sensitive forms, but this Cotter, L. A., Goldberg, M. W. and Allen, T. D. (1998). Nuclear pore approach may be limited because major alterations in pore complex disassembly and nuclear envelope breakdown during mitosis occur structure could generate viability problems. This is not the case by both nuclear envelope vesicularisation and dispersion throughout the for cell-free systems, which utilise cytoplasmic extracts from endoplasmic reticulum. Scanning 20, 250-251 amphibian eggs and demembranated sperm heads as sources Delphin, C., Guan, T., Melchior, F. and Gerace, L. (1997). RanGTP targets p97 to RanBP2, a filamentous protein localized at the cytoplasmic periphery of DNA for in vitro nucleus formation. In this system, essential of the nuclear pore complex. Mol. Biol. Cell 8, 2379-2390. nucleoporins can be depleted from extracts, which inhibits pore Feldherr, C. M. and Akin, D. (1993). Regulation of nuclear transport in formation until they are added back. Even after a 2-hour wait, proliferating and quiescent cells. Exp. Cell Res. 205, 179-186. the add back of depleted nucleoporins results in normal pore Feldherr, C. M. and Akin, D. (1994a). Variations in signal mediated nuclear transport during the cell cycle in BALB/C 3T3 cells. Exp. Cell Res. 215, formation (J. M. Cronshaw, personal communication). 206-210. Currently, however, the Xenopus system has considerably less Feldherr, C. M. and Akin, D. (1994b). The effects of SV40 large T antigen sequence information available than yeast, and the genetic and p53 on nuclear transport capacity in BALB/C 3T3 cells. Exp. Cell Res. manipulations available in yeast are not as accessible in 213, 164-171. Xenopus. However, many human nucleoporins are currently Feldherr, C. M. and Akin, D. (1995). Nuclear transport as a function of cellular activity. Membr. Prot. Trans. 2, 237-259. being used as a basis to generate Xenopus homologies, Feldherr, C. M. and Akin, D. (1997). The location of the transport gate in particularly in cases where human oncogenic transformation the nuclear pore complex. J. Cell Sci. 110, 3065-3070. produces fusion proteins that involve nucleoporins Ð at least Feldherr, C. M. and Akin, D. (1999). Signal mediated transport in the four such cases have been reported, mainly associated with amoeba. J. Cell Sci. 112, 2043-2048. Feldherr, C. M., Akin, D. and Moore, M. S. (1998). The nuclear import leukemias. Clearly any alteration in the processes affecting factor p10 regulates the functional size of the nuclear pore complex during nuclear transport will have significant effects on nuclear entry oogenesis. J. Cell. Sci. 111, 1889-1896. of transcription factors and nuclear exit of mRNAs. A further Finlay, D. R., Meier, E., Bradley, P., Horecka, J. and Forbes, D. J. (1991). The nuclear pore complex 1659

A complex of nuclear pore proteins required for pore function. J. Cell Biol. Kraemer, D. M., Strambio de Castillia, C., Blobel, G. and Rout, M. P. 114, 169-183. (1995). The essential yeast nucleoporin NUP 159 is located on the Fontoura, B., Blobel, G. and Matunis, M. J. (1999). A conserved biogenesis cytoplasmic side of the nuclear pore complex and serves as a pathway for nucleoporins: proteolytic processing of a 186kDa precursor mediated binding of transport substrate. J. Biol. Chem. 270, 19017-19021. generates Nup98 and the novel nucleoporin, Nup96. J. Cell Biol. 144, 1097- Kutay, U. Bischoff, F. R. Kostka, S. Kraft, R. Görlich, D. (1997). Export of 1112. importin alpha from the nucleus is mediated by a specific nuclear transport Fornerod, M., Boer, J., van Baal, S., Morreau, H. and Grosveld, G. (1996). factor. Cell 90, 1061-1071 Interaction of cellular proteins with the leukemia specific fusion proteins Macaulay, C. and Forbes, D. J. (1996). Assembly of the nuclear pore: DEK-CAN and SET-CAN and their normal counterpart, the nucleoporin biochemically distinct steps revealed with NEM, GTPγS and BAPTA. J. Cell CAN. Oncogene 13, 1801-1808. Biol. 132, 5-20. Fornerod, M., van Deursen, J., van Baal, S., Reynolds, A., Davis, D., Murti, Mattaj, I. W. and Engelmeier, I. (1998). Nucleocytoplasmic transport: the K. G., Fransen. J. and Grosveld, G. (1997). The human homologue of soluble phase. Annu. Rev. Biochem. 67, 265-306 yeast CRM1 is in a dynamic subcomplex with CAN/Nup214 and a novel Maul, G. G. (1977). The nuclear and the cytoplasmic pore complex: structure, nuclear pore component Nup88. EMBO J. 16, 807-816. dynamics, distribution, and evolution. Int. Rev. Cytol. 75-186. Gant, T. M., Goldberg, M. W. and Allen, T. D. (1998). Nuclear envelope Nakielny, S. and Dreyfuss, G. (1999). Transport of proteins and in and and nuclear pore assembly: analysis of assembly intermediates by electron out of the nucleus. Cell 99, 677-690. microscopy. Curr. Opin. Cell Biol. 10, 409-415. Ohno, M., Fornerod, M. and Mattaj, I. (1998). Nucleocytoplasmic transport; Goldberg, M. W. and Allen, T. D. (1993). The nuclear pore complex: three- the last 200 nanometres. Cell 92, 327-336. dimensional surface structure revealed by field emission, in-lens scanning Panté, N. and Aebi, U. (1996). Toward the molecular dissection of protein electron microscopy, with underlying structure uncovered by proteolysis. J. import into nuclei. Curr. Opin. Cell Biol. 8, 397-406. Cell Sci. 106, 261-274. Panté, N., Bastos, R., McMorrow, I., Burke, B. and Aebi, U. (1994). Goldberg, M. W. and Allen, T. D. (1995). Structural and functional Interactions and three-dimensional localization of a group of nuclear pore organisation of the nuclear envelope. Curr. Opin. Cell Biol. 7, 301-309. complex proteins. J. Cell Biol. 126, 603-617. Goldberg, M. W. and Allen, T. D. (1996). The nuclear pore complex and Powers, M. A., Forbes, D. J., Dahlberg, J. E. and Lund, E. (1997). The lamina: three-dimensional structure and interactions determined by field vertebrate GLFG nucleoporin, Nup98, is an essential component of multiple emission in-lens scanning electron microscopy. J. Mol. Biol. 257, 848-865. RNA export pathways. J. Cell Biol. 136, 241-250. Goldberg, M. W., Wiese, C., Allen, T. D. and Wilson, K. L. (1997). Dimples, Radu, A., Blobel, G. and Wozniak, R. (1993). NUP 155 is a novel nuclear pores, star-rings, and thin rings on growing nuclear envelopes: evidence for pore complex protein that contains neither repetitive sequence motifs nor structural intermediates in nuclear pore complex assembly. J. Cell Sci. 110, reacts with WGA. J. Cell Biol. 121, 1-9. 409-420. Rappsilber, J., Siniarsoglu, S., Hurt, E. C. and Mann, M. (2000). A generic Goldberg, M. W., Cronshaw, J. M., Kiseleva, E. and Allen, T. D. (1999). strategy to analyse the spatial organisation of multiprotein complexes by Nuclear pore complex dynamics and transport in higher eukaryotes. cross-linking and mass spectrometry. Anal. Chem. 72, 267-275. Protoplasma 209, 144-156. Reichelt, R., Holzenburg, A., Buhle, E. L. Jr, Jarnik, M., Engel, A. and Görlich, D., Vogel, F., Mills, A. D., Hartmann, E. and Laskey, R. A. (1995). Aebi, U. (1990). Correlation between structure and mass distribution of the Distinct functions for the two importin subunits in import. nuclear pore complex and of distinct pore complex components. J. Cell Biol. Nature 377, 246-248. 110, 883-894. Görlich, D. and Kutay, U. (1999). Transport between the and the Reipert, S., Bennion, G., Hickman, J. A. and Allen, T. D. (1999). Nucleolar cytoplasm. Annu. Rev. Cell Dev. Biol. 15, 607-660. segregation during apoptosis of haemopoietic stem cell line FDCP-Mix. Cell Gorsch, L. C., Dockendorf, T. C. and Cole, C. (1995). A conditional allelle Death Differ. 6, 334-341. of the novel repeat containing yeast nucleoporin RAT 7/NUP 159 causes Richardson, W. D., Mills, A. D., Dilworth, S. M, Laskey, R. A. and both rapid cessation of mRNA export and reversible clustering of nuclear Dingwall, C. (1988). Nuclear protein migration involves two steps: rapid pore complexes. J. Cell Biol. 129, 939-995. binding at the nuclear envelope followed by slower translocation through Grandi, P., Schlaich, N., Tekotte, H. and Hurt, E. C. (1995). Functional nuclear pores. Cell 52, 655-664. interaction of Nic96p with a core nucleoporin complex consisting of Nsp1p, Rout, M. P. and Blobel, G. (1993). Isolation of the yeast nuclear pore Nup49p and a novel protein Nup57p. EMBO J. 14, 76-87. complex. J. Cell Biol. 123, 771-783. Greber, U. F., Senior, A. and Gerace, L. (1990). A major glycoprotein of the Rout, M. P., Aitchison, J. D., Suprapto, A., Hjertaas, K., Zhao, Y. and nuclear pore complex is a membrane-spanning polypeptide with a large Chait, B. T. (2000). The yeast nuclear pore complex: Composition, lumenal domain and a small cytoplasmic tail. EMBO J. 9, 1495-1502. architecture and transport mechanism. J. Cell Biol. 148, 635-651. Greber, U. F. and Gerace, L. (1992). Nuclear protein import is inhibited by Rutherford, S. A., Goldberg, M. W. and Allen, T. D. (1997). Three- an antibody to a lumenal epitope of a nuclear pore complex glycoprotein. dimensional visualization of the route of protein import: the role of nuclear J. Cell Biol. 116, 15-30. pore complex substructure. Exp. Cell Res. 232, 146-160. Guan, T., Muller, S., Klier, G., Panté, N., Blevitt, J. M., Hauer, M., Paschal, Stoffler, D., Fahrenkrog, B. and Aebi, U. (1999a). The nuclear pore complex: B., Aebi, U. and Gerace, L. (1995). Structural analysis of the p62 complex, from molecular architecture to functional dynamics. Curr. Opin. Cell Biol. an assembly of O-linked glycoproteins that localises near the central channel 11, 391-401 of the nuclear pore complex. Mol. Biol. Cell 6, 1591-1603 Stoffler, D., Goldie, K. N., Feja, B. and Aebi, U. (1999b). Calcium-mediated Hallberg, E., Wozniak, R. W. and Blobel, G. (1993). An integral membrane structural changes of native nuclear pore complexes monitored by time-lapse protein of the pore membrane domain of the nuclear envelope contains a atomic force microscopy. J. Mol. Biol. 287, 741-752. nucleoporin-like region. J. Cell Biol. 122, 513-521. Wu, J., Matunis, M. J., Kraemer, D., Blobel, G. and Coutavas, E. (1995). Hinshaw, J. E., Carragher, B. O. and Milligan, R. A. (1992). Architecture Nup358, a cytoplasmically exposed nucleoporin with peptide repeats, Ran- and design of the nuclear pore complex. Cell 69, 1133-1141. GTP binding sites, zinc fingers, a cyclophilin A homologous domain, and a Kehlenbach, R. H., Dickmanns, A., Kehlenbach, A., Guan, T. and leucine-rich region. J. Biol. Chem. 270, 14209-14213. Gerace, L. (1999). A role for RanBP1 in the release of CRM1 from the Yang, Q., Rout, M. P. and Akey, C. W. (1998). Three-dimensional nuclear pore complex in a terminal step of nuclear export. J. Cell Biol. architecture of the isolated yeast nuclear pore complex: functional and 145, 645-657. evolutionary considerations. Mol. Cell 1, 223-234. Kiseleva, E., Goldberg, M. W., Allen, T. D. and Akey, C. W. (1998). Active Yaseen, N. R. and Blobel, G. (1997). Cloning and characterization of human nuclear pore complexes in Chironomus: visualization of transporter karyopherin-3. Proc. Nat. Acad. Sci. USA 94, 4451-4456. configurations related to mRNP export. J. Cell Sci. 111, 223-236. Yokoyama, N., Hayashi, N., Seki, T., Panté, N., Ohba, T., Nishii, K., Kuma, Kosova, B., Panté, N., Rollenhagen, C. and Hurt, E. (1999). Nup192p is a K., Hayashida, T., Miyata, T., Aebi, U., Fukui, M. and Nishimoto, T. conserved nucleoporin with a preferential location at the inner site of the (1995). A giant nuclear pore protein that binds Ran/TC4. Nature 376, 184- nuclear membrane. J. Biol. Chem. 274, 22646-22651 188.