Chromatin-Independent Nuclear Envelope Assembly Induced by Ran GTPase in Xenopus Egg Extracts

See allHide authors and affiliations

Science  26 May 2000:
Vol. 288, Issue 5470, pp. 1429-1432
DOI: 10.1126/science.288.5470.1429


The nuclear envelope (NE) forms a controlled boundary between the cytoplasm and the nucleus of eukaryotic cells. To facilitate investigation of mechanisms controlling NE assembly, we developed a cell-free system made from Xenopus laevis eggs to study the process in the absence of chromatin. NEs incorporating nuclear pores were assembled around beads coated with the guanosine triphosphatase Ran, forming pseudo-nuclei that actively imported nuclear proteins. NE assembly required the cycling of guanine nucleotides on Ran and was promoted by RCC1, a nucleotide exchange factor recruited to beads by Ran–guanosine diphosphate (Ran-GDP). Thus, concentration of Ran-GDP followed by generation of Ran-GTP is sufficient to induce NE assembly.

The NE controls access to chromatin and plays an important role in the regulation of chromosome duplication and gene expression in eukaryotic cells. In higher eukaryotes, the NE is highly dynamic in the cell cycle, being disassembled on entry into mitosis and reassembled around the segregated daughter chromosomes at telophase. The molecular mechanism of NE assembly is largely unknown, but the process can be studied in a cell-free system made fromXenopus laevis eggs (1). In this system, NE assembly around chromatin is inhibited by nonhydrolyzable guanosine 5′-triphosphate (GTP) analogues, suggesting the involvement of a GTPase (2–4). One candidate is the multifunctional GTPase Ran (5), which inSchizosaccharomyces pombe plays a role in the maintenance of NE integrity at exit from mitosis (6). Disruption of the Ran GTPase cycle by addition of regulators or dominant Ran mutants inhibits the assembly of nuclei competent for DNA replication inXenopus egg extracts (7–11) and perturbs NE structure (10, 12). However, disruption of the Ran GTPase cycle also inhibits chromatin decondensation and the establishment of nucleocytoplasmic transport during nuclear assembly (11), making it difficult to distinguish any direct role in NE formation.

To examine whether Ran plays a role in NE assembly independently of its effects on chromatin, we added glutathione-Sepharose beads coated with Ran proteins produced as fusions with glutathione-S-transferase (GST) (13) to Xenopus egg extracts (14). These fusion proteins are functional, because they produce identical effects on nuclear transport and microtubule stability as nonfusion proteins (10, 11,15). All of the beads coated with wild-type Ran-GDP became surrounded by a continuous membrane detected with the lipophilic dye 3,3′-dihexyloxacarbocyanine (DHCC) after incubation in the extracts (14), whereas beads coated with GST (Fig. 1) or the related GTPases Ras or Rab5 (16) did not attract any lipid vesicles. Transmission electron microscopy (17) showed a complete double membrane crossed by nuclear pore complexes (NPCs) assembled around Ran-GDP beads, confirming that a NE-like structure was formed (Fig. 1C). To examine the protein components of this NE, we retrieved beads from the extracts and stained nuclear pore complex proteins (nucleoporins) using the monoclonal antibody (mAb) 414 (18). Ran-GDP strongly promoted the association of nucleoporins, which were stained around the periphery of the beads, consistent with the assembly of NPCs (Fig. 2, A and B). Ran- beads also attracted the major lamin protein present in the extracts, lamin B3 (Fig. 2, C and D). This protein is actively imported and assembled into a lamina after completion of the envelope during nuclear assembly from chromatin (1), suggesting that active protein import occurs into the pseudo-nuclei formed around Ran-beads and that a lamina may be formed.

Figure 1

Sepharose beads loaded with Ran-GDP induce NE and nuclear pore complex assembly in Xenopus egg extracts. (A) Loading of proteins onto Sepharose beads. Proteins bound to 5 μl of beads loaded with glutathione-S-transferase (GST) or Ran-GDP were separated on a 10% polyacrylamide gel and stained with Coomassie blue. (B) Immunofluorescence detection of Ran (left) and lipid staining by DHCC (right) of beads incubated in Xenopus egg extracts for 120 min. DAPI was used to stain Sepharose beads; no DNA was present. The beads have a diameter of ∼100 μM, compared with 10 to 20 μm for nuclei assembled in egg extracts from sperm chromatin. (C) Transmission electron microscopy: panel 1, GST beads alone; panel 2, GST beads incubated in egg extract; panels 3 and 4 and inset, Ran-GDP beads incubated in egg extract. Nuclear pore complexes are indicated by arrows.

Figure 2

Pseudo-nuclei formed by Ran-GDP beads incorporate nucleoporins and lamins and actively import nucleoplasmin. (A) Immunofluorescence of nucleoporins detected by mAb 414. (B) Western blotting of nucleoporins detected with the same antibody. Lane 1, GST-beads alone; lane 2, GST-beads incubated inXenopus egg extract; lane 3, Ran-GDP beads incubated inXenopus egg extract; lane 4, nuclei assembled from demembranated sperm chromatin; lane 5, Xenopus egg extract. The major nucleoporins detected by mAb 414 in Xenopus nuclei are p62, p153, p210, and p340 (Ran binding protein 2). (C) Immunofluorescence and (D) Western blotting of lamin B3. Lane 1, GST-beads alone; lane 2, GST-beads incubated inXenopus egg extract; lane 3, Ran-GDP beads incubated inXenopus egg extract. Incubations were carried out for 120 min before fixing and centrifuging onto cover slips for immunofluorescence or recovery through a glycerol cushion for Western blotting. (E and F) Import of nucleoplasmin (NP). Nucleoplasmin labeled with fluorescein was added after assembly of psuedo-nuclei for 120 min, and incubation was continued for a further 60 min before fixation and recovery for fluorescence microscopy. (F) Addition of Ran proteins after formation of pseudo-nuclei but before nucleoplasmin addition.

To confirm that the NE and NPCs formed by Ran-beads were functional, we carried out an import assay using nucleoplasmin, a karyophilic protein containing a nuclear localization signal. Nucleoplasmin was taken up and strongly concentrated in the pseudo-nuclei, but not in control beads (Fig. 2E). In contrast, a fluorescent dextran too large to pass through nuclear pores was excluded from Ran-GDP beads but diffused into control beads (Fig. 4C). To determine if nucleoplasmin was being imported in a regulated manner, we added dominant Ran mutants after the completion of pseudo-nuclear assembly, but before addition of nucleoplasmin. RanQ69L (substitution of glutamine at position 69 by leucine), which is defective in GTP hydrolysis and therefore locked in the GTP-bound form (19), inhibits Ran-mediated nuclear protein import by disrupting the assembly of import complexes (20). This mutant strongly inhibited nucleoplasmin import into pseudo-nuclei assembled by using beads coated with Ran-GDP (Fig. 2F), even though NEs remained intact (16). Nucleoplasmin import was also inhibited by RanT24N (substitution of threonine at position 24 by asparagine), a mutant that is defective in nucleotide binding and probably blocks import by preventing the recycling of import factors (11). We therefore conclude that NE assembled around beads coated with Ran-GDP restricts access of macromolecules but permits the active transport of karyophilic proteins by way of the NPCs.

The nucleotide exchange factor for Ran, RCC1 (21), which binds to chromatin in a complex that may contain GTPases (22, 23), was also strongly concentrated in pseudo-nuclei (Fig. 3, A and B). The binding of RCC1 to Ran-beads was dependent on the nucleotide-bound state of Ran: Beads coated with RanQ69L-GTP bound RCC1 only poorly, whereas beads coated with RanT24N, which forms a stable inhibitory complex with RCC1 (8, 19,24), accumulated RCC1 nearly as well as wild-type Ran-beads (Fig. 3C). However, both RanQ69L-GTP and RanT24N were unable to form intact NEs, with some binding of vesicles onto the surface of beads coated with the mutants, but without fusion to form continuous membranes that could exclude fluorescent dextran (Fig. 4). Together, these results indicate that RCC1 is recruited to beads specifically by Ran-GDP. They also indicate that the exchange activity of RCC1 is required for vesicle fusion, because RanT24N blocks this step.

Figure 3

Ran-GDP recruits RCC1 to beads. (A) Immunofluorescence of RCC1. (B) Western blotting of RCC1 recovered on beads. Lane 1, GST-beads alone; lane 2, GST-beads incubated in Xenopus egg extract; lane 3, Ran-GDP beads incubated in Xenopus egg extract. (C) Immunofluorescence of RCC1 bound to beads coated with Ran-GDP, RanT24N-GDP, or RanQ69L-GTP incubated in Xenopus egg extract. Incubations were carried out for 120 min before fixing and centrifuging onto cover slips for immunofluorescence or recovery through a glycerol cushion for Western blotting.

Figure 4

NE assembly around beads requires functional Ran. (A and B) Lipid staining by DHCC is shown for beads coated with Ran-GDP, RanT24N, and RanQ69L. (A) View of whole beads. (B) Close-up showing continuous membrane staining at the surface of beads coated with Ran-GDP and vesicles at the surface of beads coated with Ran mutants. Incubations were carried out for 120 min before staining. (C) Fluorescent dextran is excluded from Ran-beads but diffuses into beads coated with RanT24N or RanQ69L. Beads were incubated for 120 min in extracts before addition of fluorescein isothiocyanate–dextran (Sigma) and incubation for a further 60 min.

To examine directly if RCC1 promotes vesicle fusion, we supplemented extracts containing Ran-GDP–coated beads with RCC1 (Fig. 5). Conversely, we added Ran binding protein 1 (RanBP1), which inhibits nucleotide exchange on Ran and stimulates the hydrolysis of Ran-GTP to Ran-GDP (25), thereby opposing the activity of RCC1. RCC1 accelerated the fusion of vesicles, forming an intact envelope after 30 min of incubation, compared with 60 min in the absence of further additions. In contrast, RanBP1 blocked vesicle fusion, preventing NE assembly even after 90 min of incubation. We conclude that generation of Ran-GTP by RCC1 from Ran-GDP concentrated on beads causes vesicle fusion to form complete NEs. The inability of RanQ69L-GTP to support vesicle fusion suggests that GTP hydrolysis on Ran is also required. This may account for the ability of nonhydrolyzable GTP analogues to block vesicle fusion (2–4).

Figure 5

Regulation of NE assembly by RanBP1 and RCC1. Ran-GDP beads were added to egg extracts, followed by 10 μM RanBP1 or 10 μM RCC1. Samples were removed at the times shown, and then stained with DAPI and DHCC; images were captured immediately.

In summary, beads coated with Ran-GDP form pseudo-nuclei with intact NEs containing functional nuclear pores, providing a potentially useful in vitro model for investigating NE formation, NPC assembly, and the establishment of nucleocytoplasmic transport. Thus, Ran plays a role in NE assembly that is distinct from its effects on chromatin structure and its role in directing nucleocytoplasmic transport. BecauseXenopus egg extracts contain Ran at a concentration of 1 to 2 μM, predominantly in the GDP-bound form (20,24), simply concentrating Ran-GDP on the surface of beads is sufficient to induce NE assembly. In somatic vertebrate cells, reassembly of the NE during telophase coincides with relocalization of Ran to chromatin (15) and stabilization of the interaction of RCC1 with chromatin (16). Consistent with these results, Ran binds to chromatin before RCC1 and before NE assembly inXenopus egg extracts (12, 15). RCC1 alone is not sufficient for NE assembly because beads coated with RCC1 do not form NEs (16). We therefore propose that concentration of Ran-GDP on the surface of chromatin in telophase promotes the binding of membrane vesicles, and then localized generation of Ran-GTP by RCC1 and subsequent GTP hydrolysis on Ran causes vesicle fusion.

Concentration of Ran-GDP promotes NE assembly, whereas Ran-GTP stabilizes microtubule asters and promotes mitotic spindle assembly inXenopus egg extracts arrested in M phase (15,26–29). A switch in the nucleotide-bound state of Ran from GTP to GDP and relocalization of Ran to chromatin may therefore coordinate NE assembly with disassembly of the mitotic spindle at the end of mitosis.


View Abstract

Navigate This Article