Self-Organization of Microtubule Asters Induced in Xenopus Egg Extracts by GTP-Bound Ran

See allHide authors and affiliations

Science  21 May 1999:
Vol. 284, Issue 5418, pp. 1356-1358
DOI: 10.1126/science.284.5418.1356


The nucleotide exchange activity of RCC1, the only known nucleotide exchange factor for Ran, a Ras-like small guanosine triphosphatase, was required for microtubule aster formation with or without demembranated sperm in Xenopus egg extracts arrested in meiosis II. Consistently, in the RCC1-depleted egg extracts, Ran guanosine triphosphate (RanGTP), but not Ran guanosine diphosphate (RanGDP), induced self-organization of microtubule asters, and the process required the activity of dynein. Thus, Ran was shown to regulate formation of the microtubule network.

Ran is an abundant Ras-like nuclear small guanosine triphosphatase (GTPase) essential for nucleus-cytosol exchange of macromolecules (1–4). Its GTPase activity is activated by RanGAP1 or Rna1p, and its nucleotide is exchanged by RCC1 (1, 2). Although all of the yeast ranmutants thus far isolated show nuclear protein import defects (3), hamster and yeast rcc1 mutants show diverse phenotypes, such as suppression of receptor-less mating process, chromosome instability, premature chromatin condensation, lack of chromosome decondensation, abnormal mRNA metabolism, and mRNA export defects (1). How Ran regulates these processes is controversial. In addition to the RanBP1 and importin β families, which are required for nucleus-cytosol exchange of macromolecules (4), Dis3p (5) and RanBPM (6) are reported to bind to RanGTP. Although Dis3p is required for ribosomal RNA processing (7), RanBPM is localized in the centrosomes and the antibodies to RanBPM inhibit microtubule aster formation in vitro (6).

The centrosome consists of a pair of centrioles surrounded by pericentriolar material (PCM) (8). Xenopusγ-tubulin ring complexes, which are made up of γ-tubulin and eight other polypeptides including Xgrip109 (9), are localized in the PCM and act as microtubule-nucleating units that cap the minus ends of microtubules (10). Although it is unknown where in the centrosome RanBPM is localized or how RanBPM participates in microtubule nucleation, the existence of RanBPM at the centrosome suggests the possibility that the Ran GTPase cycle may control microtubule organization. To address this issue, we used the in vitro microtubule assembly system of Xenopus egg extracts (11) and studied the role of Ran for microtubule organization.

The extracts were prepared from Xenopus eggs arrested in meiosis II by cytostatic factor, a cytoplasmic endogenous meiotic inhibitor (12), as described (11, 13). After incubation of the egg extracts with demembranated sperm nuclei at 24°C for 10 min, microtubule asters were formed (Fig. 1). To address the question of whether Ran participates in control of microtubule organization, we immunodepleted RCC1, which is the only known nucleotide exchange factor for Ran (1), from egg extracts (14) (Fig. 1A). More than 91% of RCC1 was removed from the egg extracts. In these RCC1-depleted egg extracts, aster formation was severely inhibited, whereas asters were formed in mock-depleted egg extracts (Fig. 1B). Consistent with this finding, RanT24N, a Ran mutant that is locked in the guanosine diphosphate (GDP)–bound form and inhibits RanGEF activity of RCC1 (15), inhibited the aster-forming action in egg extracts (Fig. 1B).

Figure 1

Requirement of RCC1 for sperm-aster formation. (A) Depletion of RCC1 from Xenopus egg extracts. Two microliters of egg extracts, untreated (lane 1), RCC1-depleted (lane 2), and mock-depleted (lane 3), was run on an SDS-polyacrylamide (10%) slab gel, transferred to a polyvinylidene difluoride membrane, and probed with antibodies to Xenopus RCC1. (B) Distribution of the number of microtubules nucleated per sperm DNA. Demembranated sperm was incubated at 24°C for 10 min with egg extracts that had been immunodepleted with antibodies toXenopus RCC1 (RCC1-depleted) or control IgG (mock-depleted) (left) or had been supplemented with RanT24N (42 μM) or buffer alone (right). More than 100 centrosomes were observed in each case. The data and error bars represent averages and standard deviations of values obtained from three different experiments.

It is possible that RCC1-depleted egg extracts failed to form asters because of loss of factors immunodepleted in association with RCC1. To exclude this possibility, we produced recombinant human RCC1 (hRCC1) inEscherichia coli and purified it (16). Purified hRCC1 was added to the RCC1-depleted egg extracts, and then mixtures were incubated with demembranated sperm nuclei. After incubation at 24°C for 10 min, microtubule asters were formed, but not in the absence of hRCC1 (Fig. 2A). Hence, RCC1 was required for microtubule aster formation in the egg extracts. Asters were also observed in the extracts where no sperm DNA was present (Fig. 2A, panel a, arrow). Indeed, when hRCC1 was added at the concentration of hRCC1, 0.22 μM, corresponding to the endogenous RCC1 concentration of egg extracts, sperm-free asters were formed (Table 1). Thus, the self-organization of microtubule asters reported previously (17) occurred in the RCC1-depleted egg extracts by addition of recombinant hRCC1.

Figure 2

Aster formation in RCC1-depleted egg extracts induced by RCC1. (A) RCC1-depleted egg extracts were preincubated for 30 min with recombinant hRCC1 (2.2 μM final concentration) (a and b) or buffer alone (c and d) and then for 10 min at 24°C with demembranated sperm nuclei and stained with mAb to α-tubulin (a and c) or with Hoechst 33342 (b and d). The arrow indicates an aster formed without sperm DNA. Scale bar, 20 μm. (B) Recombinant hRCC1 (0.22 μM final concentration) was added to RCC1-depleted egg extracts, and mixtures were incubated without sperm for 40 min at 24°C. Asters on the cover slips were stained with antibodies to Xgrip109 and NuMA as indicated. In the two right-hand panels (+mAb 70.1), the RCC1-depleted extracts were incubated with recombinant hRCC1 (0.22 μM final concentration) for 30 min and with mAb 70.1 to the dynein intermediate chain (Sigma D-5167) (final concentration of 4 μg/ml) for another 10 min at 24°C. Scale bar, 20 μm.

Table 1

Aster formation by recombinant hRCC1 and hRan. Recombinant hRCC1 or hRan bound to the indicated nucleotide was added at the indicated final concentration into the RCC1-depletedXenopus egg extracts. After incubation at 24°C for 40 min without sperm, total asters on the cover slips were counted.

View this table:

Microtubule asters induced by addition of RCC1 without sperm nuclei were stained with antibodies to Xgrip109, a Xenopus γ ring protein (9), and NuMA, a nuclear protein associated with mitotic apparatus (18), both of which are associated with microtubule minus ends. Xgrip109 and NuMA were enriched and highly focused in the poles (Fig. 2B). NuMA binds to dynein and dynactin and functions in spindle pole formation by cross-linking and sliding microtubule minus ends together (18). Hence, the presence of NuMA may indicate that the observed self-organization of microtubule asters occurred in a dynein-dependent manner. When a monoclonal antibody (mAb) to the dynein intermediate chain (mAb 70.1) (17) was added after the formation of asters, the integrity of asters was disrupted within 10 min (Fig. 2B). Thus, the dynein-dependent translocation of microtubules is apparently required for self-organization of asters induced by addition of RCC1.

To test whether the observed effect of RCC1 was dependent on its RanGEF activity, we prepared recombinant proteins of the RCC1 mutant D182, which has very little RanGEF activity (16). When recombinant hRCC1 D182 was added to RCC1-depleted egg extracts, no microtubule asters were formed (Table 1). Thus, the self-organization of microtubules into asters appears to require the RanGEF activity of RCC1. Because RCC1 catalyzes both directions of nucleotide exchange (2), we next examined a specific effect of RanGTP and RanGDP on microtubule aster formation in RCC1-depleted egg extracts. When hRanGTP was added to RCC1-depleted egg extracts, microtubule aster formation was induced in the presence of 240 μM of RanGTP, which is roughly 10 times the endogenous Ran concentration of egg extracts (Table 1). At the same concentration, RanGDP induced no aster. Thus, RanGTP, but not RanGDP, is required for microtubule aster formation.

Compared with RanGTP, hRanGTPγS induced aster formation, even in the presence of 24 μM (Table 1). Asters formed after addition of RanGTPγS, however, were very small (Fig. 3), although the number of asters formed was more than 10 times greater than that formed by the addition of RanGTP. NuMA was localized at the centers of asters induced by RanGTPγS (Fig. 3, bottom panels), revealing that microtubules were assembled into poles at the minus ends. The size of asters induced by RanGTPγS did not increase after prolonged incubation. Thus, the hydrolysis of GTP bound to Ran is essential for microtubule elongation. Neither GTP nor GTPγS alone induced microtubule organization at 240 μM. The finding that the inhibition of microtubule assembly caused by RCC1 depletion can be rescued by addition of RanGTP, but not RanGDP, indicates that the Ran GTPase cycle is involved in microtubule assembly. By analogy with Ras (19), RanGTP may carry out its function through the effectors to which RanGTP specifically binds. RanBPM is a likely effector, because it is known to bind to RanGTP and antibodies to RanBPM inhibit microtubule aster formation (6).

Figure 3

RanGTP-induced aster formation. Asters formed by addition of recombinant RanGTP (A) or RanGTPγS (B) (240 μM) to RCC1-depleted egg extracts without sperm were stained with antibodies to α-tubulin (anti–α-tubulin) (green) and NuMA (anti-NuMA) (red) as described (13). Note that asters induced by RanGTPγS were small. Scale bar, 20 μm.

The egg extracts used in this experiment were prepared from unfertilized eggs arrested in metaphase (11), and therefore no nuclear membrane was formed during the experiments. Thus, Ran affects microtubule organization independently of its role in the nucleus-cytosol exchange of macromolecules. Consistent with the notion that Ran is involved in mitotic spindle formation, RanGAP is reported to be localized in mitotic spindles (20).

  • * To whom correspondence should be addressed. E-mail: tnishi{at}


View Abstract

Stay Connected to Science

Navigate This Article