Structure, Maintenance, and Regulation of Nuclear Pore Complexes: The Gatekeepers of the Eukaryotic Genome

Abstract

In eukaryotic cells, the genetic material is segregated inside the nucleus. This compartmentalization of the genome requires a transport system that allows cells to move molecules across the nuclear envelope, the membrane-based barrier that surrounds the chromosomes. Nuclear pore complexes (NPCs) are the central component of the nuclear transport machinery. These large protein channels penetrate the nuclear envelope, creating a passage between the nucleus and the cytoplasm through which nucleocytoplasmic molecule exchange occurs. NPCs are one of the largest protein assemblies of eukaryotic cells and, in addition to their critical function in nuclear transport, these structures also play key roles in many cellular processes in a transport-independent manner. Here we will review the current knowledge of the NPC structure, the cellular mechanisms that regulate their formation and maintenance, and we will provide a brief description of a variety of processes that NPCs regulate.

A signature of eukaryotic cells is the presence of a nucleus, a membrane-based organelle that houses the cellular genome. The membrane structure that encloses the genetic material called the nuclear envelope (NE) is composed of two concentric lipid bilayers. Because the NE provides a physical barrier that completely surrounds the genome, eukaryotic cells have developed a sophisticated transport system that allows them to efficiently and selectively move molecules between the nucleus and the cytoplasm. Nucleocytoplasmic molecule exchange occurs through nuclear pore complexes (NPCs), large aqueous channels that penetrate the two membranes of the NE. As NPCs provide the only conduit into the nucleus, these large multiprotein structures are essential, and their proper formation and function is critical for cellular homeostasis. In addition to their canonical role mediating nucleocytoplasmic transport, a large amount of evidence shows that NPCs and their components play multiple transport-independent functions. Moreover, the expression levels of NPC components can vary significantly among different cell types and tissues, and the residence time at NPCs is widely different between different components. These findings have exposed that NPCs are dynamic in nature, can change their composition or stoichiometry in different cell types, and can regulate a multitude of cellular processes either in a transport-dependent or independent fashion. Our goal here is to provide an overview of the latest advances in the understanding of NPC formation, dynamics, maintenance, and functions.

NPC STRUCTURE

In the 1950s, Callan and Tomlin used electron microscopy (EM) of amphibian oocyte nuclei to provide the first evidence for the existence of pores at the NE (Fig. 1; Callan and Tomlin 1950). With these initial images of the nuclear membranes, the authors suggested for the first time that the rims of the pores in the outer nuclear membrane, now known as the outer scaffold ring, were built up above the nuclear membrane. Seventeen years later, Joseph Gall provided the first description of the eightfold rotational symmetry of NPCs (Fig. 1; Gall 1967). A multitude of subsequent studies using more powerful imaging and structural analysis techniques helped to further understand the three-dimensional structure of NPCs over the past half century (Fig. 1; Hinshaw et al. 1992; Akey and Radermacher 1993; Stoffler et al. 2003; Beck et al. 2004, 2007; Maco et al. 2006; Lim et al. 2008; Elad et al. 2009; Frenkiel-Krispin et al. 2010; Maimon et al. 2012; Szymborska et al. 2013; Löschberger et al. 2014; Eibauer et al. 2015; von Appen et al. 2015; Schwartz 2016; von Appen and Beck 2016; Beck and Hurt 2017; Kim et al. 2018; Mosalaganti et al. 2018; Vallotton et al. 2019; Zhang et al. 2020). But because NPCs are such large protein complexes, dissecting the architecture of these channels at the molecular level represented a challenging undertaking, and it was not until more recently that the exact position of nucleoporins within the NPC structure was determined with high resolution. The recent crystallization of many NPC components, the breakthrough generation of high-resolution images of the NPC structure by cryo-electron tomography, and the determination of nucleoporin interactions have allowed several groups to develop highly accurate models of the NPC molecular organization, in particular of the NPC scaffold (Fig. 1; Alber et al. 2007; Bui et al. 2013; Stuwe et al. 2015; von Appen et al. 2015; Kosinski et al. 2016; Lin et al. 2016; Kim et al. 2018; Allegretti et al. 2020). The detailed structural organization of the NPC and its components has been extensively reviewed (Kabachinski and Schwartz 2015; Hoelz et al. 2016; Beck and Hurt 2017; Hampoelz et al. 2019a; Lin and Hoelz 2019) and will be described here concisely.

Figure 1.

Timeline depicting hallmark structural findings of the nuclear pore complex (NPC). For a better illustration of key findings, images were obtained from original publications or provided by the authors. (Image of first electron microscopy [EM] of NPCs by Callan and Tomlin 1950, reprinted with permission from The Royal Society © 1950. Original high-resolution scanning images kindly provided by Dr. Martin Goldberg [Ris 1989; Goldberg and Allen 1992]. Cryo-electron tomography of NPCs, image reprinted from Beck et al. 2004 with permission from the American Association for the Advancement of Science [AAAS] © 2004. Atomic models of the NPC core, image reprinted from Lin et al. 2016 and Kosinski et al. 2016 with permission from AAAS © 2016.)

It is important to note that the massive NPCs are highly modular complexes built by the repetition of ∼30 different proteins called nucleoporins (the exact numbers varies slightly in different species) (Table 1; Rout et al. 2000; Cronshaw et al. 2002; Obado et al. 2016; Field and Rout 2019). NPCs are channels of eightfold rotational symmetry that sit on pores generated at the NE by the fusion of the inner and outer nuclear membranes. These channels consist of a central scaffold built by an inner ring embedded in the NE and two outer rings. The scaffold has eight identical spokes and twofold symmetry parallel to the NE plane and surrounds the NPC central transport channel, which is filled with nucleoporins rich in phenylglycine (FG) repeats (GLFG, FxFG, PxFG, or SxFG). Through weak hydrophobic interactions between their FG domains, central channel nucleoporins set the diffusion barrier of the NPC (Ribbeck and Görlich 2002; Frey et al. 2006; Frey and Görlich 2007, 2009; Xu and Powers 2013; Schmidt and Görlich 2015; Zahn et al. 2016; Fisher et al. 2018; Gu et al. 2019; Celetti et al. 2020; Dormann 2020), interact with transport receptors to mediate the nuclear transport (Iovine et al. 1995; Bayliss et al. 2000, 2002; Sträßer et al. 2000; Fribourg et al. 2001; Isgro and Schulten 2007a,b; Terry and Wente 2007; Zahn et al. 2016; Tan et al. 2018; Hayama et al. 2019), and help to stabilize the structure of the NPC (Onischenko et al. 2017). Finally, specific nucleoporins associate on either side of the symmetric core to form the cytoplasmic filaments and the nuclear basket structures.

Table 1.

Nuclear pore complex (NPC) composition in budding yeast and vertebrates^a

At the molecular level, and because of the eightfold symmetry of the NPC, nucleoporins are present in eight or multiples of eight copies (Ori et al. 2013; Kim et al. 2018; Rajoo et al. 2018). While the outer rings of the NPC are mainly formed by copies of the Nup107-160 complex, also known as the Y-complex, the inner ring is formed by the repetition of the Nup93-Nup205 complex (recently, for review, see Lin and Hoelz 2019). In humans, the inner and outer rings are sporadically linked through a component of the Nup93-Nup205 complex, nucleoporin Nup155 (Kosinski et al. 2016; Kim et al. 2018; Lin and Hoelz 2019). The main nucleoporins that fill the central channel are the components of the Nup62 complex (Nup62, Nup54, and Nup58) (Finlay et al. 1991; Strawn et al. 2004; Sharma et al. 2015). These three nucleoporins are anchored to the inner ring of the NPC through Nup93 and extend their disordered FG-containing domains into the central channel (Sachdev et al. 2012; Fischer et al. 2015; Stuwe et al. 2015). The Nup62 complex together with Nup98, an FG nucleoporin that associates with the cytoplasmic and nuclear sides of the NPC (Griffis et al. 2003), define the selective permeability of the NPC (Frey and Görlich 2007; Ader et al. 2010). Notably, high concentrations of Nup98 FG repeats have been shown to undergo a phase separation in vitro and to form a hydrogel that recapitulates the permeability and transport properties of the NPC (Frey et al. 2006; Frey and Görlich 2007; Hülsmann et al. 2012). These findings support the idea that phase transitions induced by the interactions of FG repeats might govern the permeability barrier of the NPC (Schmidt and Görlich 2015, 2016; Fisher et al. 2018; Celetti et al. 2020; Dormann 2020). In this model, nuclear transport receptors can move through the molecular condensate barrier by outcompeting the nucleoporin FG–FG interactions (Ribbeck and Görlich 2001; Antonin 2013; Zahn et al. 2016). Even though a molecular condensate environment represents a highly likely scenario for the central channel of the NPC, the intrinsically disordered nature of the FG-containing regions of nucleoporins have prevented their structural definition using conventional techniques, and several other possibilities have also been proposed to explain the permeability barrier of the NPC (D'Angelo and Hetzer 2008; Mincer and Simon 2011; Li et al. 2016; Hayama et al. 2017).

In metazoans, the cytoplasmic side of the NPC is built by the recruitment of Nup358, Nup214, and Nup88, Nup358 being the main component of the cytoplasmic filaments. This protein is not present in budding yeast, Saccharomyces cerevisiae (Kim et al. 2018), which might explain why these cells show shorter (or different) filament structures (Kiseleva et al. 2004). The nuclear side of the NPC is mainly composed of the nucleoporin translocated promoter region (TPR), and its associated proteins Nup153 and Nup50, which form a nuclear basket-like structure tethered to the outer nuclear ring. This nuclear basket consists of eight filaments and a distal intranuclear ring.

Although the overall structure of the NPC is evolutionarily conserved, differences between species do exist, size being the most obvious variance. For example, while the human NPC core is 120 nM wide by 80 nM in height, and ∼110 MDa (Kosinski et al. 2016), its S. cerevisiae counterpart is 95 nM by 60 nM and 52 MDa (Kim et al. 2018). A difference in size is partly explained by the outer scaffold rings. While the human NPC outer rings have 16 copies of the Y complex, the S. cerevisiae outer rings only have eight (Bui et al. 2013; Ori et al. 2013; von Appen et al. 2015; Rajoo et al. 2018). This different stoichiometry is a result of humans having double outer rings while yeast has single ones. Interestingly, the Chlamydomonas reinhardtii alga has an “in-between” structure with a single cytoplasmic outer ring like yeast, but a double nuclear outer ring like humans (Mosalaganti et al. 2018). A recent study also identified that S. cerevisiae NPCs lack the extra molecules Nup157 or Nup170 (homologs of human Nup155) connecting the inner to the outer rings and do not a show Nup188 or Nup192 (homolog of human Nup205) in the outer rings (Kim et al. 2018), as described for the human NPC (Kosinski et al. 2016). This challenges a previous model suggesting that core scaffolds were identical between both species (Lin et al. 2016).

Although NPCs are membrane-embedded channels, only three to four nucleoporins are transmembrane proteins and all the rest are soluble (Table 1; Wozniak and Blobel 1992; Hallberg et al. 1993; Wozniak et al. 1994; Chial et al. 1998; Miao et al. 2006; Stavru et al. 2006; Chadrin et al. 2010). But some soluble nucleoporins have amphipathic lipid-packing sensor (ALPS) domains, which are sequences 20–40 amino acids long containing hydrophobic residues that can bind to curved lipid bilayers, providing additional links to the nuclear membranes (Drin et al. 2007; Leksa et al. 2009; Doucet et al. 2010; Mitchell et al. 2010; Kim et al. 2014; Mészáros et al. 2015; Lin et al. 2016; Lin and Hoelz 2019; Nordeen et al. 2020). Even though the structure and domains of nucleoporins seem to be conserved among many species, these proteins show poor sequence conservation, which has made it difficult to identify them in some organisms. Moreover, some nucleoporins are restricted to specific organisms. For example, the cytoplasmic filament protein Nup358/RanBP2 has so far only been described in metazoans, while the transmembrane nucleoporin Pom121 is restricted to vertebrates (Funakoshi et al. 2007, 2011; Kim et al. 2018). With a few exceptions, nucleoporins have a remarkably limited set of domains mostly limited to structure and protein–protein interactions, such as α-solenoids, β-propellers, coiled-coiled motifs, WD40, and Zn-finger domains (see Lin and Hoelz 2019 for details).

At first sight, the NPC looks like a modular structure of very simple composition built by the repetition of roughly 30 different proteins (Rout et al. 2000; Cronshaw et al. 2002). But increasing evidence points toward a much higher complexity for this structure. First, several nucleoporins have differential expression among cell types and tissues, and NPCs of different composition and stoichiometry of nucleoporins have been reported, indicating that NPCs can vary among different cell types (D'Angelo et al. 2012; Ori et al. 2013, 2016). Moreover, recent evidence exposed that NPCs also change during specific cellular processes and in response to different cellular conditions (Asally et al. 2011; D'Angelo et al. 2012; Raices and D'Angelo 2012; Rodriguez-Bravo et al. 2018; Liu et al. 2019). Adding to this complexity, many nucleoporins show multiple splicing isoforms, most of which have not yet been characterized (Capitanchik et al. 2018; Hampoelz et al. 2019a). Although some of these splice variants could work “off pore,” as recently described for Pom121 (Franks et al. 2016), it is possible that some might be used to change the properties or functions of NPCs. A significant number of nucleoporins also have short residence time at NPCs, meaning that they are constantly exchanged, and further highlighting the dynamic nature of these structures (Rabut et al. 2004). Additionally, multiple proteomic studies have shown that most nucleoporins have a significant number of posttranslational modifications, including phosphorylation, ubiquitination, N- and O-glycosylation, sumoylation, and acetylation (for review, see Hampoelz et al. 2019a). The physiological relevance of most of these modifications is still unknown; some of these modifications might be used to alter the structure, dynamics, and functions of NPCs. Consistent with this idea, phosphorylation of nucleoporins was shown to modulate their interaction with transport receptors (Kosako and Imamoto 2010; Wigington et al. 2020), O-glycosylation modifies NPC integrity, permeability, and nuclear transport (Mizuguchi-Hata et al. 2013; Zhu et al. 2016; Eustice et al. 2017), acetylation of NPC components regulates cell-cycle entry and asymmetric cell division (Kumar et al. 2018), and ubiquitylation and sumoylation of nuclear basket nucleoporins modulate their association, transcriptional regulation, nuclear migration, and the cellular response to stress (Hayakawa et al. 2012; Texari and Stutz 2015; Niño et al. 2016; Folz et al. 2019).

NPC LIFE CYCLE IN DIVIDING CELLS

Proliferating cells have a constant need for NPC assembly. When cells divide, they split their contents into two daughter cells that receive only half the NPCs of the mother. Therefore, these cells need to assemble a new set of channels for the next round of cell division to avoid a progressive dilution of NPC numbers. In organisms undergoing closed mitosis, like S. cerevisiae and Schizosaccharomyces pombe, cell division was long considered to occur with no NE or NPC breakdown, meaning that these cells only assemble new NPCs into an intact NE (Boettcher and Barral 2013). Interestingly, evidence for a local disassembly of NPCs and the breakdown/remodeling of the NE in the bridge that connects the segregating S. pombe daughter nuclei were recently provided (Dey et al. 2020; Expósito-Serrano et al. 2020). Although these findings do not indicate a second mechanism of NPC formation in these cells, they provide evidence for similarities in the process of NE remodeling between open and closed mitosis.

In metazoans, mitosis requires a total breakdown of the NE including the full disassembly of NPCs (open mitosis) (Anderson and Hetzer 2008). Consequently, daughter cells need to reassemble their NPCs by recycling maternal components when the NEs reform during M-phase (postmitotic NPC assembly), and they also need to assemble NPCs into an intact NE later during interphase to double their number before the next cell division (interphase NPC assembly). Although mitotic and interphase NPC assembly result in the formation of indistinguishable structures, significant evidence indicates that these processes are fundamentally different (Otsuka and Ellenberg 2018).

In metazoans, NPC disassembly during mitosis is triggered by massive phosphorylation of nucleoporins by mitotic kinases, mostly cyclin-dependent kinase 1 (CDK1), Polo-Like Kinase 1 (PLK1), and the NIMA-related kinase (NEK) family (Macaulay et al. 1995; Favreau et al. 1996; Miller et al. 1999; De Souza et al. 2004; Onischenko et al. 2005; Glavy et al. 2007; Lusk et al. 2007; Rajanala et al. 2014; Linder et al. 2017; de Castro et al. 2018). Mitotic phosphorylation of nucleoporins induces the dissociation of protein complexes and leads to their release from the NE. NPC disassembly occurs synchronously and through a stepwise process that is faster than NPC formation, and does not follow the exact reverse order (Dultz et al. 2008). In mammalian cells, hyperphosphorylation of Nup98 has been found to be an early and rate-limiting step in NPC disassembly (Laurell et al. 2011; Linder et al. 2017). Phosphorylation of this nucleoporin results in its loss from NPCs, followed by the loss of peripheral nucleoporins and disruption of the nuclear permeability barrier (Linder et al. 2017). Phosphorylation of Nup35 (also referred to as Nup53) is another early event in NPC disassembly that leads to its dissociation from Nup155 and NDC1, and to the destabilization of nuclear pores (Linder et al. 2017). Similar stepwise disassembly mechanisms have been observed in other organisms (Kiseleva et al. 2001; Cotter et al. 2007; Katsani et al. 2008).

As mitosis progresses and the NEs of daughter cells begin to reform, so do NPCs. Mitotic NPC assembly is considerably faster than interphase assembly, taking minutes versus approximately an hour; synchronized, meaning that hundreds of NPCs form at the same time; and critical for the establishment of the nuclear permeability barrier (D'Angelo et al. 2006; Dultz et al. 2008; Dultz and Ellenberg 2010; Otsuka et al. 2016, 2018; Onischenko et al. 2020). Mitotic NPC assembly also occurs in a mitotic environment, where the cytoplasmic and nuclear contents are mixed. Reassembly is triggered by dephosphorylation of nucleoporins (Antonin et al. 2008; Huguet et al. 2019), and NPCs are built in a stepwise process with components inherited from the mother cell (Fig. 2A; Maul 1977; Burke and Ellenberg 2002; Dultz et al. 2008). In the mitotic cytoplasm, many nucleoporins are bound to importin β, an association that blocks the interaction between these proteins and prevents NPC reassembly (Zhang et al. 2002; Harel et al. 2003; Walther et al. 2003b). In the vicinity of chromatin, importin β-nucleoporin complexes encounter high concentrations of RanGTP, which is generated by Ran's DNA-bound guanine-exchange factor RCC1 (Ohtsubo et al. 1989; Li et al. 2003). RanGTP binding to importin β releases nucleoporins allowing them to initiate NPC formation on chromosomes (Harel et al. 2003; Walther et al. 2003b; Franz et al. 2007; Rotem et al. 2009). A similar role for the export receptors CRM1, exportin-t, and exportin-5 in nucleoporin sequestration and release during postmitotic NPC assembly has recently been reported (Nord et al. 2020). Evidence from several laboratories indicate that the process starts during anaphase with the recruitment of the Y complex nucleoporin Elys to decondensing chromatin (Fig. 2A; Rasala et al. 2006, 2008; Franz et al. 2007; Gillespie et al. 2007; Dultz et al. 2008; Inoue and Zhang 2014). This seeding step is followed by the sequential association of the scaffold Nup107-160 complex, the membrane protein Pom121, components of the Nup93-Nup205 complex, and later, components of the central channel, and other nucleoporins (Fig. 2A; Bodoor et al. 1999; Belgareh et al. 2001; Daigle et al. 2001; Walther et al. 2003a,b; Antonin et al. 2005; Franz et al. 2005, 2007; Rasala et al. 2006, 2008; Gillespie et al. 2007; Dultz et al. 2008; Theisen et al. 2008; Mitchell et al. 2010; Sachdev et al. 2012; Vollmer et al. 2012; Eisenhardt et al. 2014). The stepwise recruitment of nucleoporins to chromatin results in the formation of intermediate structures, known as pre-pores, that were initially observed by EM in Xenopus oocytes and Drosophila embryos, and later confirmed in mammalian cells (Sheehan et al. 1988; Macaulay and Forbes 1996; Goldberg et al. 1997; Drummond et al. 2006; Otsuka et al. 2018). Using correlative single-cell live imaging and high-resolution scanning electron tomography, a recent study shed light into the mechanisms of postmitotic NPC formation. This work shows that during NE reformation, nuclear membranes form from highly fenestrated endoplasmic reticulum membrane sheets, and NPCs assemble into small preexisting membrane holes that progressively dilate as the nuclear pore rings and the central channel mature (Fig. 2A; Otsuka et al. 2018). This progressive dilation of membrane openings explains some of the early structural intermediates observed using a Xenopus in vitro nuclear assembly system (Goldberg et al. 1997). On the other hand, the formation of NPCs into preexisting membrane holes of the fenestrated ER membranes contradicts previous findings, suggesting that the formation of the NE membranes precedes NPC assembly (Macaulay and Forbes 1996; Fichtman et al. 2010; Lu et al. 2011).

Figure 2.

Stepwise representation of different mechanisms of nuclear pore complex (NPC) assembly. (A) Postmitotic NPC assembly. (B) Interphase NPC assembly. (ONM) Outer nuclear membrane, (INM) inner nuclear membrane.

Once the daughter nuclei have formed, cells start assembling new NPCs into an intact NE. Even though this is known as interphase NPC assembly, the process starts during telophase and occurs continuously until G2 (Winey et al. 1997; Maeshima et al. 2006; Otsuka and Ellenberg 2018). Interphase NPC assembly is sporadic and requires the fusion of the inner and outer nuclear membranes. Like mitotic NPC assembly, this is also a stepwise process, although it neither requires Elys nor follows the exact same order of nucleoporin recruitment (Fig. 2B; Doucet et al. 2010; Dultz and Ellenberg 2010; Onischenko et al. 2020). Although interphase NPC assembly requires importin β, RanGTP, and the Nup107-160 complex from the cytoplasmic and nuclear sides of the NE (D'Angelo et al. 2006), a recent study exposed that interphase NPC assembly does not occur symmetrically from both sides of the NE but rather through an inside-out mechanism (Otsuka et al. 2016). Interphase NPC assembly is believed to start with the recruitment of the Nup107-160 complex to the inner nuclear membrane by Nup153 (Vollmer et al. 2015) and requires the Pom121 transmembrane nucleoporin and the Sun1 inner nuclear membrane protein (Talamas and Hetzer 2011). The recruited nucleoporins start to form a dome-shaped intermediate structure that progressively grows in diameter and depth, pushing the inner membrane inward until it contacts the outer nuclear membrane (Fig. 2B). Strikingly, the early mushroom-shaped intermediate structure already shows an eightfold rotational symmetry and the presence of Nup107, suggesting that the outer ring is one of the first structures to assemble. Once the growing structure pushes the nuclear membranes in close proximity, they fuse, and the process continues with the recruitment of additional nucleoporins, such as the Nup358, to form the mature NPC structure (Otsuka et al. 2016). Several factors required for NPC assembly into an intact NE, such as karyopherin β (importin β), RanGTP, and key nucleoporins, were originally identified in S. cerevisiae (Vasu and Forbes 2001; Lusk et al. 2002; Ryan et al. 2003, 2007; Madrid et al. 2006), indicating conservation of this assembly mechanism across species.

It is important to note that NPC formation is not restricted to the NE, and stacks of nuclear pores in ER membrane sheets, known as annulate lamellae, can be observed during embryogenesis and in many cancer cells (Maul 1970; Dabauvalle et al. 1991; Cordes et al. 1995, 1996). In Drosophila embryos, annulate lamellae NPCs (AL-NPCs) were found to lack several components of the central channel, cytoplasmic filaments, and nuclear basket, suggesting that they mostly consist of the core structure (Fig. 3; Hampoelz et al. 2016). These partially assembled structures were found to be added to the NE during the nuclei expansion that occurs during the fast cell-cycle divisions in early embryogenesis (Fig. 3), suggesting that these structures represent a stock of partially assembled NPCs that are used to support fast proliferation (Hampoelz et al. 2016). Even though the assembly of annulate lamellae NPCs also requires Ran and CRM1, the partially assembled channels were found to form from nucleoporin condensates (Hampoelz et al. 2016, 2019b). Notably, granules of Nup358, a nucleoporin recruited at the late stages of interphase NPC assembly, were identified as the main drivers of NPC formation in the ER. Moreover, AL-NPCs do not contain Nup153, a key player in NPC assembly into the NE membranes (Vollmer et al. 2015; Hampoelz et al. 2016). These findings suggest that NPC assembly in annulate lamellae might employ a mechanism different from the mitotic and interphase assembly processes.

Figure 3.

Annulate lamellae contribution to nuclear pore complex (NPC) assembly during embryogenesis. Individual components involved in the processes are shown in the figure. (ER) Endoplasmic reticulum, (ONM) outer nuclear membrane, (INM) inner nuclear membrane.

NPC MAINTENANCE AND TURNOVER

The building of the massive NPC machinery is a complicated process that can sporadically fail leading to defective or nonfunctional structures that can affect cellular physiology. Recent studies in yeasts uncovered a novel surveillance mechanism used by cells to identify and eliminate aberrantly assembled NPCs (Webster et al. 2014, 2016; Thaller et al. 2019). In this mechanism, the inner nuclear membrane proteins Heh1 and Heh2 recruit CHM7 and members of the ESCRT complex to defective NPC assembly intermediates. These proteins work in concert to seal defective NPCs, as well as destabilize and eliminate the aberrant structures. Disrupting the function of this surveillance machinery leads to the accumulation of defective assembly intermediates, affects nucleocytoplasmic transport, and results in the loss of nuclear compartmentalization (Webster et al. 2014, 2016; Thaller et al. 2019).

Selective autophagy of NPCs, termed NPC-phagy, a mechanism that might contribute to turnover nuclear pores, was recently identified in S. cerevisiae. This process requires the autophagy core component ATG11 and is mediated by the interaction of the autophagy marker protein ATG8 with Nup159, the homolog of human Nup214 (Allegretti et al. 2020; Lee et al. 2020; Tomioka et al. 2020). Nucleoporin mutant strains that either interrupt NPC assembly or lead to the aggregation of mature NPCs at the NE suggest that selective autophagy might be required for degradation of fully assembled NPCs rather than elimination of stalled NPC assembly intermediates (Allegretti et al. 2020; Lee et al. 2020; Tomioka et al. 2020). This selectivity is based on the fact that mature NPCs are exposed to the cytoplasm and can contact the autophagy machinery while aberrant assembly intermediates are sealed and hidden from it. These intermediates are covered by membranes, either because their formation stalled before the inner and outer nuclear membranes fused, or because they were sealed by the ESCRT-dependent surveillance mechanism (Webster et al. 2016).

In nondividing cells, the dynamics and maintenance of NPCs is different. First, as cells exit the cell cycle, they shut down the expression of essential nucleoporins and nuclear pore formation is strongly down-regulated (D'Angelo et al. 2009). Whereas in quiescent cells some level of NPC assembly and turnover is maintained (Toyama et al. 2019), in differentiated postmitotic cells the NPC scaffolds are one of the longest-lived protein complexes, and can be maintained for months, maybe even years, without turnover (D'Angelo et al. 2009; Savas et al. 2012; Toyama et al. 2013). In these cells, the NPC scaffolds seem to be renewed by the gradual replacement of individual subunits rather than the replacement of the entire structure (Toyama et al. 2013). Whether autophagy contributes to NPC turnover or to the elimination of malfunctioning transport channels in differentiated cells has not yet been investigated.

NPC FUNCTIONS

Because NPCs are the sole connection between the nucleus and the cytoplasm, their canonical function has traditionally been considered the regulation of nucleocytoplasmic transport. Although this is still recognized as the main function of NPCs, these structures are also one the largest protein assemblies of cells, and a great amount of evidence indicates that they play many other cellular roles. Because of space limitations, we will provide a summarized description of the NPC roles, which by no means represents a comprehensive account of the extensive literature in the field.

As gatekeepers of the nucleus, NPCs play two main functions: (1) they establish the permeability barrier of the NE, and (2) they work in coordination with transport receptors to ferry molecules in and out of the nucleus. As described above, both functions are mediated mostly by the FG nucleoporins of the central channel, although components of the cytoplasmic filaments and nuclear basket also play a role in nucleocytoplasmic transport (Bastos et al. 1996; Ullman et al. 1999; Walther et al. 2001; Forler et al. 2004; Makise et al. 2012). Ions and small molecules diffuse freely through NPCs, but larger molecules need to be actively transported through these channels. There is not an exact diffusion limit for NPCs, but a range between 30 and 60 kDa is largely accepted. Interestingly, several reports show that a permeability limit of NPCs can change in response to different cellular conditions during the cell cycle, as cells age, and in disease (Feldherr and Akin 1990; Roehrig et al. 2003; Belov et al. 2004; Shahin et al. 2005; D'Angelo et al. 2009; Dultz et al. 2009; Eftekharzadeh et al. 2018). In addition, transport receptors were recently reported to contribute to the permeability barrier of these channels (Lowe et al. 2015; Barbato et al. 2020). Although a few proteins have been found to move through NPCs by direct interaction with nucleoporins (Fagotto et al. 1998; Yokoya et al. 1999; Xu et al. 2003; Tsuji et al. 2007; Wagstaff and Jans 2009), the majority of the transported cargoes require transport receptors (Mosammaparast and Pemberton 2004; Kimura and Imamoto 2014). These are molecules that recognize nuclear localization signals (NLS), in the case of importins, or nuclear export signals (NES), in the case of exportins, in the cargo molecules and ferry them through NPCs. The family of nuclear transport receptors (collectively known as karyopherins or importin β family members) encompasses >20 different importins and exportins (Mosammaparast and Pemberton 2004; Kimura and Imamoto 2014). Specific exportins can also bind and export RNAs including microRNAs, ribosomal RNAs, tRNAs, and small nuclear RNAs, but the nuclear export of the majority of mRNAs is carried out through a completely different mechanism. The processes of protein and RNA nuclear transport have been the subject of several reviews (Köhler and Hurt 2007; Carmody and Wente 2009; Wente and Rout 2010; Cautain et al. 2015; Okamura et al. 2015; Scott et al. 2019) and will not be described in detail here.

But NPCs are far more than just the doors of the nucleus. Evidence accumulated over the past few decades, and particularly in the last one, have exposed that NPCs act as scaffolds for the regulation of a myriad of cellular process in a transport-independent manner. For example, a large amount of data shows that NPCs play a direct role in the regulation of gene expression and chromatin organization. They do this by recruiting transcription factors and chromatin modulators to the nuclear periphery to regulate local enhancers, gene activity, and chromatin state by promoting gene looping, and by modulating gene positioning (for reviews, see Randise-Hinchliff and Brickner 2016; Buchwalter et al. 2019; Pascual-Garcia and Capelson 2019; Sun et al. 2019). Nucleoporins have also been shown to regulate gene expression “off pore” (in the nuclear interior) by interacting and modulating the activity of chromatin and transcriptional regulators (Capelson et al. 2010b; Kalverda et al. 2010; Liang et al. 2013; Pascual-Garcia et al. 2014; Jacinto et al. 2015; Franks et al. 2016). There is clear evidence that NPCs also play a key role in the response to DNA damage and replication stress (Bukata et al. 2013; Freudenreich and Su 2016; Duheron et al. 2017; Mackay et al. 2017; Rodriguez-Berriguete et al. 2018; Gaillard et al. 2019; Horigome et al. 2019; Aguilera et al. 2020; Pinzaru et al. 2020), and contribute to RNA processing and surveillance (Lewis et al. 2007; Ikegami and Lieb 2013; Bonnet and Palancade 2015).

Several protein-modifying enzymes, including histone deacetylases (Kehat et al. 2011), SUMO proteases (Panse et al. 2003; Chow et al. 2012, 2014), poly(ADP-ribose) polymerases (PARPs) (Meyer-Ficca et al. 2015; Carter-O'Connell et al. 2016; Kirby et al. 2018), phosphatases (Sales Gil et al. 2018; Wigington et al. 2020), and kinases (Faustino et al. 2011; Martino et al. 2017) have been found associated with NPCs, and nucleoporin Nup358 has been shown to be an E3 SUMO ligase important for the sumoylation of many cellular factors (Pichler et al. 2004). Even proteosomes have been recently found to be associated with NPCs (Albert et al. 2017). These findings indicate that NPCs also act as hubs for signal transduction, posttranslational modifications, and potentially localized protein degradation. Cell-cycle regulatory proteins Mad1 and Mad2 associate with NPCs during interphase (Campbell et al. 2001; Iouk et al. 2002), and the presence of Mad1 at the pores has been shown to be important for proper cell-cycle timing (Rodriguez-Bravo et al. 2014; Jackman et al. 2020). In addition, several nucleoporins have been shown to localize to kinetochores and/or the mitotic spindle during mitosis where they contribute to regulate their functions (Belgareh et al. 2001; Stukenberg and Macara 2003; Joseph et al. 2004; Loïodice et al. 2004; Blower et al. 2005; Galy et al. 2006; Rasala et al. 2006; Schetter et al. 2006; Zuccolo et al. 2007; Lussi et al. 2010; Mishra et al. 2010; Schweizer et al. 2013). These findings directly link NPCs and nucleoporins to cell-cycle regulation. Examples of additional functions that nucleoporins have been shown to play off-pore include a role for Nup188 in centriole duplication (Vishnoi et al. 2020), Nup358 function in cytoskeletal organization (Joseph and Dasso 2008), the regulation of mTORC1 activity by Seh1 and Sec13 being part of the GATOR2 complex (Shaw 2013), and of course the regulation of gene expression and chromatin organization as described above (Capelson et al. 2010b; Kalverda et al. 2010; Liang et al. 2013; Pascual-Garcia et al. 2014; Jacinto et al. 2015; Franks et al. 2016).

CONCLUDING REMARKS

In the past two decades there has been an explosion in our understanding of NPCs. We now know where almost every nucleoporin sits within the NPC structure, how specific nucleoporins build the permeability barrier and regulate nuclear transport, and how these complex structures are formed and maintained in dividing and differentiated cells. We also have learned that NPCs play many different functions, independent of their role in nucleocytoplasmic transport. Because these functions span several different fields—some examples are the recently identified roles of NPCs in the development and function of immune, muscle, and brain cells (for reviews, see Guglielmi et al. 2020); their contribution to cancer, heart disease, and neurodegeneration (Simon and Rout 2014; Beck et al. 2017; Jühlen and Fahrenkrog 2018; Hutten and Dormann 2019; Burdine et al. 2020); and their key roles in gene expression regulation, and the maintenance of genome integrity (Capelson et al. 2010a; Bukata et al. 2013; Raices and D'Angelo 2017; Sun et al. 2019)—investigators with a variety of backgrounds have become interested in the biology of NPCs, and it can be expected that our knowledge of these structures will continue to expand significantly in the coming decade.