Ran is an abundant Ras-like GTPase that was initially characterized for its role in modulating the nucleo-cytoplasmic transport of macromolecules across the nuclear envelope (NE). However, it is also critical for other cellular processes, including mitotic spindle assembly and post-mitotic nuclear envelope assembly. Several excellent reviews detail its role in these diverse processes (Gruss and Vernos, 2004; Harel and Forbes, 2004; Hetzer et al., 2002; Macara, 2001). Here, I attempt to give an overview of Ran's functions and reconcile these with the possibility that a common molecular mechanism underlies them. Current models for how Ran modulates different cellular processes center on the biochemical concept that activation factors exist in inhibitory complexes with transport receptors (e.g. importin β), which are relieved and activated by RanGTP in a spatiotemporal manner.
Ran, like other GTPases, switches between a GTP-bound and a GDP-bound form. These transitions are regulated by a guanine nucleotide exchange factor (GEF) termed RCC1 (regulator of chromosome condensation 1) and the GTPase-activating protein RanGAP1. The hydrolysis by RanGAP1 requires an additional factor, RanBP1, which possesses a RanGTP-binding domain. In interphase, where Ran's function in nucleo-cytoplasmic transport has been extensively studied, these regulators of Ran are asymmetrically distributed: RCC1 is chromatin-bound and hence nuclear, whereas RanGAP1 and RanBP1 are cytoplasmic. In metazoans, RanGAP1 is covalently modified with a small ubiquitin-like peptide, SUMO1, which targets it to the cytoplasmic face of the nuclear pore complex (NPC) through association with the nuclear pore protein RanBP2 (also called Nup358) (Mahajan et al., 1997). RanBP2 has four RanBP1-like domains and is therefore believed to function similarly to RanBP1, although in vivo evidence of this is lacking.
The asymmetric localization of the regulators is predicted to create a RanGTP gradient across the NE: RanGTP levels being higher in the nucleus and lower in the cytoplasm. This asymmetry directs nucleo-cytoplasmic transport by regulating the assembly and disassembly of import and export complexes (reviewed by Gorlich and Kutay, 1999; Macara, 2001). The transport of macromolecules across the NE is generally mediated by a family of RanGTP-binding receptors, termed `karyopherins'. These include importin β and CRM1 (chromosome region maintenance 1), which mediate the import and export of a characteristic set of cargos containing nuclear localization signals (NLSs) and nuclear export signals (NESs), respectively. In the cytoplasm, where the RanGTP level is low, the import complex is formed when NLS-containing cargo is recognized by importin α and assembled with importin β. The import complex travels through the NPC into the nucleus, where it disassembles following binding of importin β to the abundant nuclear RanGTP. Conversely, NES-bearing export cargo binds CRM1 and RanGTP in the nucleus and translocates through the NPC. In the cytoplasm, RanGTP hydrolysis by RanGAP1 and RanBP1/RanBP2 disassembles the complex, releasing the export cargo. The RanGDP resulting from GTP hydrolysis in the cytoplasm is recycled back to the nucleus by the RanGDP-binding protein NTF2 (nuclear transport factor 2). RanGTP thus acts as a positional cue defining the nuclear compartment, and directs the disassembly and assembly of import and export complexes, respectively. Importin β cycles back to the cytoplasm in a RanGTP-dependent manner, whereas importin α requires an additional factor, CAS (cellular apoptosis susceptibility protein), for its recycling. Both CRM1 and NTF2 can relocate to the nucleus and cytoplasm, respectively, independently of RanGTP.
Mitotic spindles are essential for the proper organization and precise segregation of chromosomes. A comprehensive picture of spindle assembly is still lacking, and two potential mechanisms have been proposed (reviewed by Karsenti and Vernos, 2001). In one, the duplicated centrosomes positioned at opposite poles nucleate dynamic microtubules that are captured and stabilized by kinetochores and this leads to formation of a bipolar spindle. In the other, chromosomes are the source of spindle organization and play a critical role nucleating and stabilizing microtubules. This is proposed to lead to the convergence of these microtubules to form proper bipolar spindles as a consequence of a crosslinking activity generated by microtubule motors and other factors. Both mechanisms probably operate in vivo, contributing to different extents in different systems. For example, in soluble extracts of Xenopus eggs, which arrest in the second metaphase of meiosis, bipolar spindles assemble in the absence of centrosomes and kinetochores, on artificial chromosomes (Heald et al., 1996). In this instance, the chromosomes mediate spindle formation entirely through the second mechanism. However, in somatic cells, centrosomes and kinetochores play critical roles.
Experimental considerable evidence shows variations in the localization of Ran regulators/mediators and the effect of the Ran pathway on spindle assembly in Xenopus compared with somatic mammalian cells (Di Fiore et al., 2004; Quimby and Dasso, 2003). Here they are therefore treated in two different sections. However, it is possible that the fundamental mechanism that Ran uses for spindle assembly in both the systems is similar.
Spindle assembly in Xenopus egg extracts
Although Ran cycle mutants in yeast show defects in mitosis, this was originally thought to be due to defective nucleo-cytoplasmic transport of mitotic regulators (Sazer and Dasso, 2000). Recent studies of Xenopus egg extracts, however, have confirmed that Ran plays a role in mitotic spindle assembly independently of its role in nuclear transport (reviewed by Kahana and Cleveland, 1999). Further studies also showed that importin β is a downstream effector of Ran in spindle assembly and a model has emerged that is consistent with its action in nuclear transport. According to this, after nuclear envelope breakdown at the onset of mitosis, the chromatin-bound RCC1 continues to concentrate RanGTP around chromosomes, and RanGAP1 and RanBP1/2 hydrolyze GTP on Ran away from chromatin, giving rise to a concentration gradient of RanGTP. Spindle assembly factors (SAFs) that mediate microtubule stabilization and spindle assembly, such as TPX2, NuMA and XCTK2, are proposed to be sequestered in inhibitory complexes with importin β away from chromatin. Near chromatin, however, RanGTP could bind and displace importin β by a mechanism similar to that involved in nuclear transport, releasing and activating the SAFs around chromosomes (Dasso, 2001). TPX2 has been shown to have a role targeting Aurora A, a critical mitotic kinase, to the spindle and activating it (Tsai et al., 2003). In addition, RanGTP has been shown to regulate Eg5 - a microtubule-crosslinking motor protein required for the formation of a bipolar spindle (Wilde et al., 2001) - probably by activating it indirectly through Aurora-A-mediated phosphorylation. Recent work has identified additional SAFs that include Rae1 (RNA export 1 protein), NuSAP (nucleolar and spindle-associated protein), lamin B and HURP (hepatoma up-regulated protein) (Blower et al., 2005; Koffa et al., 2006; Ribbeck et al., 2006; Sillje et al., 2006; Tsai et al., 2006).
Spindle assembly in somatic cells
The Ran pathway appears to regulate spindle assembly in somatic cells by regulating at least two processes: centrosomal integrity/cohesion and kinetochore-microtubule interactions. Ran localizes to centrosomes in somatic cells and when RanGTP levels are decreased by overexpression of RanBP1, centrosomal cohesion is lost (Di Fiore et al., 2003; Keryer et al., 2003). Importin β is probably involved because its overexpression results in supernumerary spindle poles, possibly owing to inappropriate centrosome duplication (Ciciarello et al., 2004). The results are consistent with a model in which hypothetical centrosomal integrity factors exist in inhibitory complexes with importin β and get released by RanGTP at centrosomes to maintain cohesion (Harel et al., 2003). However, further studies are needed to identify such factors and validate this model.
RanBP2 and RanGAP1 associate with kinetochores in mammalian somatic cells during mitosis and play an important role in mediating proper kinetochore-microtubule interactions (Joseph et al., 2002; Joseph et al., 2004). Furthermore, the kinetochore localization of the RanBP2-RanGAP1 complex is regulated by RanGTP through the export receptor CRM1 (Arnaoutov et al., 2005). How CRM1 modulates the kinetochore localization of RanBP2-RanGAP1 and what is the mechanism by which this complex regulates microtubule-kinetochore interactions require further investigation. The RanBP2-RanGAP1 complex may regulate receptor-cargo interactions by influencing the levels of RanGTP. Alternatively, RanBP2 may itself act as an effector for RanGTP, by recruiting some additional factor to the kinetochore or by acting enzymatically on a target localized there (Arnaoutov et al., 2005). In either case, RanGTP appears to monitor proper kinetochore-microtubule interactions and thereby precise chromosome segregation in these cells.
Post-mitotic nuclear envelope assembly
In higher eukaryotes, at the end of mitosis the NE is reformed around the segregated DNA. This probably involves recruitment of membrane vesicles around the chromatin, their fusion and assembly of NPCs on the NE. Initial studies in Xenopus egg extract showed that both RanGTP and importin β function in this process (reviewed by Clarke and Zhang, 2001). The relative levels of importin β and RanGTP seem to be important for proper membrane fusion and nuclear pore assembly. Whereas RanGTP stimulates membrane fusion and nuclear pore assembly, importin β negatively regulates these events (Harel et al., 2003; Walther et al., 2003). However, the negative effect of importin β on pore assembly cannot be overcome by RanGTP, which indicates that additional factor(s) are involved (Harel et al., 2003). These findings are again consistent with the notion that a hypothetical regulatory factor(s) is present in inhibitory complexes with importin β, and is released by RanGTP. In the case of NPC assembly, the association of importin β with nucleoporins, such as Nup107, Nup153 and Nup358, can be relieved by RanGTP, which suggests that these proteins act as NPC assembly factors (Walther et al., 2003). Future studies should reveal other factors regulated by RanGTP and the details of the mechanism involved.
Studies of tsBN2, a mutant BHK (baby hamster kidney) cell line that possesses a temperature-sensitive RCC1 allele, and Xenopus egg extracts, revealed that the Ran pathway also regulates entry into mitosis (reviewed by Sazer and Dasso, 2000). At the restrictive temperature, tsBN2 cells bypass the DNA-replication (S phase) checkpoint that monitors the completion of DNA replication before the onset of M phase, which results in precocious entry into mitosis, NE breakdown and premature chromosome condensation. In Xenopus extracts, by contrast, reducing RanGTP levels leads to inappropriate activation of this checkpoint. Lowering RanGTP levels thus has opposite effects on the DNA-replication checkpoint in somatic cells and Xenopus extracts. Nevertheless, the data clearly indicate that Ran regulates entry into mitosis. Future work is required if we are to understand the underlying mechanisms and explain these differences.
In addition, recent work indicates that Ran regulates the spindle assembly checkpoint, a signaling pathway that ensures that the cells pause in metaphase until all the duplicated chromosomes are properly attached to microtubules from opposite poles (Arnaoutov and Dasso, 2003). The checkpoint operates partly by recruiting a set of checkpoint regulators to kinetochores (reviewed by Musacchio and Hardwick, 2002). These are removed from kinetochores when the spindle checkpoint is inactivated, triggering the onset of anaphase. Elevated levels of RanGTP in Xenopus extracts, achieved by addition of RCC1, lead to the removal of checkpoint regulators from kinetochores and inactivation of the spindle checkpoint. When RanGAP1 and RanBP1 are depleted from the extract, the spindle checkpoint is similarly inactivated. By contrast, addition of RanGAP1 and RanBP1 to extracts with exogenous RCC1 restores the spindle checkpoint activation. The Ran pathway thus appears to play a role in regulation of the spindle checkpoint. Importin β appears to be dispensable for this, however (reviewed by Li et al., 2003). Future investigations should reveal how Ran regulates these checkpoints that monitor entry into mitosis and the metaphase-anaphase transition.
The Ran GTPase appears to be a master regulator and coordinator of events that require intimate crosstalk between chromatin (or the nucleus) and the cytoplasm. Ran seems to use similar underlying mechanisms in mitosis, post-mitotic NE formation, and nuclear transport-key players being members of the karyopherin family of proteins. Because RCC1 is predominantly chromatin-bound throughout the cell cycle, RanGTP levels can be elevated near chromatin. RanGTP could thus act as a spatial marker describing where the chromatin or nucleus is. This information seems to be made use of by cells for nuclear transport in interphase and coordination of various other events during mitosis. In general, the concept that transport factors sequester different activators and inhibit their function until relieved by RanGTP in a spatial manner (near chromatin) appears to underlie the mechanism by which Ran regulates various cellular events. Efforts should now focus on understanding the processes regulated by Ran in detail, defining all the players involved and the molecular mechanisms by which the different events are precisely orchestrated during the cell cycle.
I thank Mary Dasso for the useful comments on this manuscript, and apologize to the many researchers whose work could not be cited owing to space constraints.