When Overexpressed, a Novel Centrosomal Protein, RanBPM, Causes Ectopic Microtubule Nucleation Similar to γ-Tubulin

The HeLa cell line is derived from human uterine cervical carcinoma, and MRC5 is a primary human cell culture. COS7 cells expressing the SV-40 early gene are derived from a green monkey cell line CV1 (Gluzman, 1981). CHO-k1 cells are derived from the Chinese hamster. Cells were cultured in Dulbecco's modified Eagle's medium (DME) (NIPRO Hymedium; Sigma Chemical Co., St. Louis, MO) supplemented with 10% fetal bovine serum (Sigma Chemical Co.) in a humidified atmosphere containing 10% CO₂.

Cells on glass coverslips cultured in dishes (35-mm-diam) were transfected with 1 μg of plasmid DNA using the Lipofectamine protocol (GIBCO BRL, Gaithersburg, MD). The efficiency of transfection was usually ∼5%.

1.5 kb of human RanBPM-ORF carried in the plasmid containing RanBPM cDNA cloned was amplified by PCR using as the 5′ primer: 5′ AAGGTCGACACATGAATAGACTACCAGGTTGG 3′, and as the 3′ primer: 5′ CGCAAGCTTTTCAAATCAGCAGAGCTAGTC 3′. The resultant DNA fragment was digested with the restriction enzymes SalI and HindIII, and inserted into the Sal1/HindIII sites of pET28, resulting in pET28–RanBPM, which express RanBPM tagged with T7 at the NH₂ terminus.

pcDEB-T7–RanBPM: 1.7 kb of T7-RanBPM was excised from pET28– RanBPM with the restriction enzymes XbaI and HindIII, then inserted into the XbaI and HindIII sites of the pcDEB vector (Hayakawa et al., 1990) which had been digested with XbaI and HindIII enzymes, resulting in pcDEB-T7–RanBPM.

pAS404–RanBPM: 1.5 kb of RanBPM containing the open reading frame (ORF) was amplified by PCR using as the primers, 5′ AGGGTCGACCATGAATAGACTACCAGGTTGG 3′, and 5′ CGCAAGCTTTTCAAATCAGCAGAGCTAGTC 3′, digested with the HincII enzyme, and then inserted into the Sma1 site of the pAS404 vector derived from pAS1 (Durfee et al., 1993).

pACTII–RanG19V/pACTII–RanT24N: 1.3 kb of the fragments containing Ran were excised with the restriction enzymes NcoI and BamHI from pET8c–RanG19V and pET8c–RanT24N (Dasso et al., 1994), and inserted into the NcoI and BamHI sites of pACTII (Durfee et al., 1993).

To amplify the 5′ end of RanBPM cDNA, three kinds of the primers consisting of the RanBPM nucleotides 186–212, 219–245, or 907–936 were prepared based on the nucleotide sequence of RanBPM (GenBank/ EMBL/DDBJ accession number AB008515). Using these primers, human cDNAs were amplified either from mRNA-isolated HeLa cells by AmpliFINDER RACE kit (Clontech, Palo Alto, CA) and then by TaKaRa EX Taq (TaKaRa, Otsu City, Shiga, Japan) or from Human Burkitt Lymphoma cDNA library with Marathon-Ready cDNA kit (Clontech) and then by TaKaRa LA Taq (TaKaRa).

RNA filters were prepared as described previously (Yokoyama et al., 1995) and prehybridized with 100 μg/ml of salmon sperm DNA at 42°C for 2 h in buffer containing 0.5% SDS, 50% formamide, 5× SSPE (0.15 M NaCl, 10 mM NaH₂PO₄, pH 7.4, 1 mM EDTA), 5× Denhardt's solution, and then incubated with ³²P-labeled cDNA for 24–48 h. After hybridization, the filters were washed in the following manner: twice in 2× SSPE plus 0.5% SDS for 30 min at room temperature, twice in 0.1× SSPE plus 0.5% SDS for 30 min at 70°C , and then once in 2× SSPE at room temperature. Finally, filters were dried and subjected to autoradiography.

The antibodies to RanBPM, against the peptide FDIEDYMREWRTKIQ were prepared in the rabbit as described (Nakashima et al., 1993) and then affinity purified by using antigenic peptide-coupled Sepharose columns.

Cells on coverslips were washed for 15 s at 37°C with microtubule stabilization buffer (0.1 M Pipes, pH 6.9, 1 mM EGTA, 4 M glycerol, and 1 mM GTP) as described (Shu and Joshi, 1995), incubated for 1 min at 37°C in the same buffer containing 0.5% Triton X-100, rinsed with microtubule stabilization buffer, and then plunged into methanol at −20°C for 5 min. After fixation, cells were rehydrated in PBS and doubly stained with the primary antibodies; the mAb to T7-tag (Novagen, Madison, WI), α-tubulin (N356; Amersham Pharmacia Biotech, Piscataway, NJ) or γ-tubulin (T6557; Sigma Chemical Co.), and with rabbit affinity-purified anti– γ-tubulin antibodies (Masuda et al., 1996) or RanBPM. After staining, the antibodies were diluted with PBS-T (PBS containing 0.1% Tween 20) for 1 h at room temperature. After rinsing with PBS-T, cells were stained for 45 min at room temperature with the secondary antibodies, fluorescein isothioyanate (FITC)-conjugated goat anti–mouse IgG (AMI3408; Biosource) or anti-rabbit IgG (ALI3408; Biosource, Camarillo, CA), Texas red-conjugated goat anti–rabbit IgG (55675; Cappel, Malvern, PA), or anti-mouse IgG (N2031; Amersham). Cells were finally stained with Hoechst 33342 and then mounted on Vectashield (Vector, Burlingame, CA).

Zeiss Axio Photo (Carl Zeiss Inc., Thornwood, NY) was used with the standard microscopic method. Digital imaging of stained cells was obtained using the laser-scanning microscope, Zeiss LSM 310 (Carl Zeiss Inc.) or Olympus LSM-GB200 system (Tokyo, Japan), and printed by pictrography 3000 (Fujix Tokyo, Japan) through Adobe Photoshop^TM 3.0J (Adobe Systems Inc., San Jose, CA).

Isolation of centrosomes from HeLa cells and CHO cells was carried out in accordance with previous method (Bornens et al., 1987) but with some minor modifications. Cultured HeLa or CHO cells were incubated with 10 μg/ml nocodazole and 5 μg/ml cytochalasin B for 2 h, rinsed with isolation buffer (1 mM Tris, pH 8, 0.5 mM EGTA, 0.1% β-mercaptoethanol), and then lysed by swaying the dishes in isolation buffer containing 0.5% NP-40, at 4°C for 10 min. Next, a one-fiftieth volume of PE buffer (0.5 M Pipes, pH 7.2, 0.1 M EGTA) was added to the extract which was then subjected to the discontinuous sucrose density gradient set in an SRP28-SA tube (Hitachi, Tokyo, Japan) with 3.5 ml of 60% sucrose (wt/wt), 3.5 ml of 40% sucrose (wt/wt) prepared in 20 mM Pipes, pH 6.8, 0.5 mM MgCl₂, 1 mM EGTA and 0.1% β-mercaptoethanol, and run at 14,500 rpm for 1 h. Fractions were collected from the bottom and analyzed for microtubule nucleation ability as follows, according to Mitchison and Kirschner (1984).

1 μl of each fraction was incubated with 12.5 μl of tubulin solution (40 mM tubulin of porcine brain, 80 mM Pipes, pH 6.8, 1 mM MgCl₂, 1 mM EGTA and 1 mM GTP) at 37°C for 10 min. Microtubules were fixed by adding glutaraldehyde, sedimented onto poly-l-lysine–coated coverslips, and then subjected to immunofluorescence. Using Zeiss Axio Photo (Carl Zeiss Inc.), the aster possessing more than 10 microtubules was counted. We counted the total number of the asters formed on a coverslip three times. The mean value and the standard deviation (SD) of the obtained numbers are shown in the text.

E. coli-produced wild-type and mutated Ran proteins were prepared as described (Dasso et al., 1994).

Cells were lysed in buffer containing 62.5 mM Tris-HCl, pH 6.8, 100 mM dithiothreitol, 2% (wt/vol) SDS, and 10% glycerol. Cellular proteins were electrophoresed in an SDS-12% polyacrylamide slab gel, transferred to a polyvinylene difluoride (PVDF) membrane and probed with antibodies as described previously (Nakashima et al., 1993).

The human cDNA clone B8 was isolated by the two-hybrid method using human Ran as a bait (Yokoyama et al., 1995). Northern analysis using the B8 clone as a probe revealed the presence of a 3.1-kb mRNA in human HeLa cells (Fig. 1,A). Subsequently, the human cDNA library was screened using the cDNA clone B8 as the probe. Finally, the cDNA clone of 2.8 kb, whose nucleotide sequence has been deposited as GenBank/EMBL/DDBJ accession number AB008515, was isolated. Since there was no stop codon at the site upstream from the first methionine, we repeatedly carried out 5′ RACE using as primers the nucleotides localized either near the 5′ end or the middle of the RanBPM cDNA. But there was no cDNA clones extended from the 5′ end of RanBPM cDNA. We concluded the ORF of RanBPM (Fig. 1 B) to consist of 500 amino acid residues, encoding a protein with a calculated molecular mass of 55,079 Da.

To identify the protein encoded by the isolated cDNA clone, the antibody was prepared against the peptide of the putative ORF highlighted in Fig. 1,B, shaded box. The obtained serum was purified by the affinity column to the peptide used as the antigen. Total cell extracts were prepared from cultures of human HeLa and MRC5 cell lines, and then subjected to immunoblotting analysis using the affinity-purified antibodies. In both cell extracts, a major band of protein with a molecular mass of 55 kD that is consistent with the molecular mass calculated based on the putative ORF, was specifically recognized (Fig. 1,C). The higher bands of 57 kD which were also specifically recognized by the antibodies (Fig. 1 C, compare lanes 1 with 3), could in fact be modified forms of the protein encoded by the putative ORF or other proteins cross-reactive to the antibodies. Although the absence of in-frame termination codons upstream allows for the possibility of a small extension at the NH₂ terminus, we concluded that the putative ORF encodes a predicted protein of 55 kD, designated as RanBPM.

By homology search, several mouse cDNA fragments, the accession numbers of which are shown in Fig. 1,B, were found to encode a section of RanBPM. The amino acid sequences deduced from these mouse cDNA fragments are 100% identical to those deduced from human RanBPM, indicating that RanBPM is well conserved among mammalians. To confirm this issue, the library of hamster cDNA was screened for RanBPM cDNA. The isolated cDNA fragment encoded a polypeptide that is identical to human RanBPM (Fig. 1 B, dotted arrows).

By BLAST search, we found that the COOH-terminal half of the S. cerevisiae ORF; YGL227w, is highly homologous to the human RanBPM. Percent amino acid identity of RanBPM with YGL227w is 29.8% and the probability that such sequence similarity is realized by chance is less than 10⁻¹³, when calculated as described (Toh et al., 1983). In particular, the amino acid sequence of RanBPM (1–78) is homologous to the SPRY domain found in several proteins, although its function remains unknown (Schultz and Bork, 1997).

Immunoblotting analysis using the affinity-purified anti-RanBPM antibodies showed that whereas several protein bands were recognized in the extracts of HeLa cells, only a single band of 55 kD was recognized in the extracts of MRC5 cells. To determine the cellular localization of RanBPM, cultures of MRC5 cells were doubly stained with the affinity-purified anti-RanBPM antibodies (red) and the mAb to α-tubulin (green) (Fig. 2 A). When both staining patterns overlapped, the microtubule was found to be nucleated from the matrix which was stained with the affinity-purified anti-RanBPM antibodies. These results suggested that the protein encoded by the B8 clone was localized within the centrosome, and thereby RanBPM stands for Ran-binding protein in MTOC.

To confirm that RanBPM was localized in the centrosome, cultures of MRC5 cells were doubly stained by the affinity-purified anti-RanBPM antibodies (red) and by the mAb to γ-tubulin (green) (Fig. 2,B). When superimposed, the color changed to yellow, thereby revealing that RanBPM was colocalized with γ-tubulin, one of the important centrosomal residents (Oakley and Oakley, 1989; Kalt and Schliwa, 1993; Kellogg et al., 1994). To confirm that the γ-tubulin stained in MRC5 cells was located within the centrosome, cells were double stained by the affinity-purified γ-tubulin antibodies (red) (Masuda et al., 1996) and the mAb to α-tubulin (green) (Fig. 2 C). When superimposed, the affinity-purified γ-tubulin antibodies stained the central matrix of the microtubule network, as expected. In mitotic HeLa cells, the area corresponding to the centrosome was also stained by the affinity-purified anti-RanBPM antibodies (data not shown).

Based on these staining results, we concluded that RanBPM was localized within the centrosome throughout the cell cycle, similar to γ-tubulin.

To determine the biological function of RanBPM, we overexpressed RanBPM cDNA in COS cells and examined its effect on the microtubule network. T7-fused RanBPM cDNA carried on the pcDEB vector was introduced into COS cells as described in Materials and Methods. 36 h later, transfected cells were fixed and doubly stained by the affinity-purified anti-RanBPM antibodies (red) and by the mAb to α-tubulin (green). In contrast to cells transfected with the vector alone (Fig. 3,A, panel c), a normal radial network of microtubules was broken in cells transfected with T7-RanBPM cDNA (the representative figures are shown in Fig. 3,A, panels a and b). When transfected cells were doubly stained by the mAb to the T7-tag (green) and the affinity purified anti–γ-tubulin antibodies (red), RanBPM and γ-tubulin were found to be distributed as spots throughout the cytoplasm (Fig. 3,B). Both staining spots were colocalized when superimposed (Fig. 3,B, superimposition), indicating that there is some type of interaction between γ-tubulin and RanBPM. We thought that upon overexpression of T7-RanBPM, γ-tubulin was recruited onto RanBPM, resulting in reorganization of the microtubule network similar to the case of overexpressed γ-tubulin (Shu and Joshi, 1995). To confirm this issue, we monitored the recovery of microtubules after complete disassembly induced by lowering the temperature to 0°C as described (Joshi et al., 1992; Shu and Joshi, 1995). RanBPM cDNA or, as a control, the vector alone, was transfected into COS cells. To avoid rapid cell death, we transfected a smaller amount of RanBPM cDNA per cell (10 ng/dish), compared with the experiment described above (1 μg/dish). Under this condition, the average number of RanBPM-staining spots was ∼3–5 per cell. 24 h later, transfected cells were placed on ice for 1 h. Cells were then incubated in fresh medium at 30°C for varying time periods ranging from 0 to 5 min, lysed to remove free tubulin and then fixed to visualize the initiation of microtubule assembly sites in the green channel and RanBPM in the red channel by double immunofluorescence microscopy (Fig. 4). In cells transfected by the vector alone, after return to 30°C, short microtubules emerged from a single RanBPM-stained spot and became progressively elongated as previously reported (Shu and Joshi, 1995). In contrast to cells transfected by the vector alone, multiple RanBPM-stained spots appeared in cells transfected with T7-tagged RanBPM cDNA (Fig. 4). After incubation at 30°C for 1 min, α-tubulin gathered around multiple RanBPM-stained spots. Subsequently, short microtubules emerged from the multiple RanBPM-stained spots, although the microtubules were shorter than the microtubules that emerged from the centrosome of the untransfected cells. The number of ectopically nucleated microtubules was the same as the number of RanBPM-stained spots. Thus, we concluded that overexpressed T7-tagged RanBPM caused ectopic nucleation of the microtubule assembly in vivo.

We then determined the relationship between RanBPM and microtubule nucleation, using the isolated centrosome. To achieve this, the centrosome extracts were prepared from HeLa cells by sucrose-density gradient centrifugation as described (Mitchison and Kirschner, 1984; Bornes et al., 1987). Fractions containing the centrosomes were determined by microtubule nucleation ability (Fig. 5,A). Immunoblotting analysis revealed that both RanBPM and γ-tubulin were cosedimented with the centrosome fractions (Fig. 5,B). The protein bands higher than 55 kD which were fractionated into the top fractions were recognized by the affinity-purified anti-RanBPM antibodies. The major band of these corresponds to the band of 57 kD recognized in the total extract of HeLa cells (Fig. 1 C). We do not know whether they are modified forms of RanBPM or proteins cross-reactive to the affinity-purified anti-RanBPM antibodies as mentioned above. These proteins were concentrated into the top fractions, suggesting that they were localized in the cytoplasm. Similarly, a majority of γ-tubulin was fractionated into the top fraction, being consistent with the previous report that the majority of γ-tubulin is localized in the cytoplasm (Stearns and Kirschner, 1994).

To determine the localization of RanBPM in the microtubule asters assembled in vitro, the asters were doubly stained by the mAb to α-tubulin (green) and by the affinity-purified anti-RanBPM antibodies (red). As expected from the in vivo results, RanBPM was localized at the central part of the microtubule asters (Fig. 6) where γ-tubulin was also localized (see Fig. 9). These results are consistent with the notion that RanBPM is one of the centrosomal components which is involved in microtubule nucleation. To address this issue, the centrosome fractions were preincubated with the affinity-purified anti-RanBPM antibodies, and then assayed for aster formation. Although the preimmune IgG had no effect on aster formation, the number of microtubules nucleated by the centrosome was greatly reduced by addition of anti-RanBPM antibodies (Fig. 7). In a control reaction mixture containing the preimmune IgG or the buffer alone, the total number of asters formed on a coverslip was 1442 (SD = 27.0) and 1489 (SD = 51.1), respectively. In contrast, it was 151 (SD = 10.1) in the presence of anti-RanBPM antibodies. Thus, the aster-forming ability of the centrosome factions was reduced to about 10% of the buffer alone by the addition of the affinity-purified anti-RanBPM antibodies. Under the same conditions, the polymerization of microtubules without centrosome fractions was not inhibited by the affinity-purified anti-RanBPM antibodies (data not shown), this being consistent with the result showing that the length of the microtubules was not reduced by the addition of anti-RanBPM antibodies (Fig. 7).

Human RanBPM cDNA was isolated by the two-hybrid method using human Ran as a bait (Yokoyama et al., 1995). To confirm whether or not RanBPM interacts specifically with GTP-Ran, mutant forms of Ran, G19V-Ran, and T24N-Ran were prepared as described (Dasso et al., 1994; Kornbluth et al., 1994). The amino acid residues 19 and 24 of Ran were conserved between ras and Ran. By analogy of ras, G19V-Ran is thought to be locked in an activated state through inability to hydrolyze GTP (McGrath et al., 1984) and T24N-Ran is thought to remain predominantly in a GDP-bound form since the corresponding mutation in ras^H have a profoundly decreased affinity for GTP in vitro (Feig and Cooper, 1988). Indeed, G19V-Ran and T24N-Ran preferentially binds to GTP and GDP, respectively (Kornbluth et al., 1994; Lounsbury et al., 1996).

The cDNA of G19V-Ran and T24N-Ran, both of which were fused in frame with the GAL4-activation domain of pACT (Durfee et al., 1993), were introduced into cultures of the strain Y190 [pAS404-RanBPM]. Transformants were selected in synthetic medium lacking leucine and tryptophan, and plated on synthetic medium either lacking histidine, tryptophan, and leucine but containing 10 mM of 3-aminotriazole (+) or lacking tryptophan, leucine and 3-aminotriazole (−). In the presence of 3-aminotriazole, the strain Y190 [pAS404-RanBPM, pACT-G19VRan] papillated, whereas the strain Y190 [pAS404-RanBPM, pACT-T24NRan] did not (Fig. 8). This result indicates that RanBPM specifically interacts with GTP-Ran.

Based on the above results, we then determined the effects of Ran on microtubule aster formation. The centrosome fractions were preincubated either with GTP-, GTPγS-, or GDP-bound Ran, or with G19V-Ran, and were then assayed for aster formation by the addition of tubulin. The total number of asters formed on a coverslip was 702 (SD = 69.2) without the Ran preparation. By addition of either GTPγS-Ran or G19V, it was reduced to 53 (SD = 7.5) and 56 (SD = 5.2), respectively. On the other hand, in the presence of GTP-Ran or GDP-Ran, the total number of asters was 650 (SD = 26.0) and 592 (SD = 3.7), respectively. Thus, the ability of the centrosome to nucleate microtubules was significantly reduced by addition of GTPγS-bound Ran and G19V-Ran as shown in Fig. 9. By immunoblotting analysis, we found that the centrosome fractions contained a considerable amount of RanGAP1, but no RCC1 (data not shown). Therefore, GTPγS-bound Ran, which can not be hydrolyzed, should remain stable in the reaction mixture.

In the presence of GTPγS-Ran, the central part of the microtubule asters which was stained by both the affinity-purified anti-RanBPM antibodies and the mAb to γ-tubulin significantly faded, compared with the case of buffer alone (Fig. 9). However, the length of the microtubules was not reduced, and the nonhydrolyzable GTP-Ran seemed to inhibit the nucleation of microtubule assembly, similar to the anti-RanBPM antibodies. Indeed, the microtubule elongation was not inhibited by GTPγS-Ran or G19V-Ran in the absence of the centrosome factions (data not shown). It is notable that the central part of the microtubule asters did not fade when the anti-RanBPM antibodies were added (Fig. 7). The mechanism for inhibiting microtubule–aster formation, therefore, seemed to be different between the anti-RanBPM antibodies and the nonhydrolyzable GTP-Ran.

RanBPM was most frequently isolated by two-hybrid screening of the human cDNA library using human Ran as a bait (Yokoyama et al., 1995). 38 out of a total of 80 clones isolated encode RanBPM. In the same screening, several fragments of RanBP2 cDNA which encode one of four Ran-binding domains of RanBP2 were isolated, revealing that our screening of Ran-binding proteins was functional. In this context, it is highly likely that RanBPM is yet another Ran-binding protein, although RanBPM has no Ran-binding domain similar to either that of RanBP1 (Coutavas et al., 1993) or that of importin β (Görlich et al., 1997).

The colocalization of RanBPM with γ-tubulin indicates that RanBPM is localized within the centrosome. The γ-TuRC purified from Xenopus egg extracts contains at least seven different proteins (Zheng et al., 1995). The proteins homologous to S. cerevisiae SPB components Spc97p and Spc98p have been identified in the γ-TuRC and are localized in the centrosome (Martin et al., 1998; Murphy et al., 1998; Tassin et al., 1998). Recently, Wigge et al. (1998) prepared a highly enriched spindle pole preparation and identified a total of 12 known and 11 novel components of the SPB. Among these, there was found no protein homologous to RanBPM, although we found a possible RanBPM homologue of S. cerevisiae ORF; YGL227w by Blast search. It remains to be investigated whether RanBPM directly interacts with γ-tubulin. However, as we discuss below, we demonstrated a functional interaction between RanBPM and γ-tubulin.

Overexpression of the cloned RanBPM cDNA causes both a reorganization of the microtubule network and an ectopic microtubule nucleation. These results indicate that the cloned RanBPM cDNA has the biological activity of nucleating the microtubule assembly in vivo. The finding that overexpressed RanBPM cDNA causes ectopic microtubule nucleation is quite similar to the case of γ-tubulin reported previously (Shu and Joshi, 1995). Of great interest is that newly formed RanBPM spots were costained with the mAb to γ-tubulin. Although the mechanism for the ectopic microtubule nucleation has not been clarified, it may be possible that the overexpression of RanBPM activates γ-TuRC or recruits γ-tubulin to the centrosome scaffolds, resulting in microtubule nucleation.

The in vitro microtubule nucleation is inhibited not only by the addition of anti-RanBPM antibodies, but also by the addition of nonhydrolyzable GTP-Ran. The central part of the microtubule asters stained either by the mAb to γ-tubulin or by anti-RanBPM antibodies fades in the presence of nonhydrolyzable GTPγS-Ran or a dominant GTP-bound mutant of Ran, G19V-Ran (Kornbluth et al., 1994; Lounsbury et al., 1996). This is in contrast to the finding that upon addition of the anti-RanBPM antibodies, the central part of microtubule asters does not change in size. The anti-RanBPM antibodies may structurally hinder the microtubule assembly at the nucleation site. In contrast, it is unclear how the nonhydrolyzable GTP-Ran inhibits the nucleation of microtubule assembly. Recently, it has been reported that the higher-order organization of microtubule-nucleating sites is represented by a centrosomal lattice containing pericentrin and γ-tubulin (Dictenberg et al., 1998). Therefore, one of possibilities is that the dynamic stability of the centrosomal lattice is regulated by Ran through RanBPM. The tight binding of nonhydrolyzable GTP-Ran to RanBPM may make the centrosomal lattice unstable, resulting in a shrinking of the central part. However, it could be also possible that GTPγS-Ran inhibits the aster formation independently from RanBPM. It is a very interesting question how Ran regulates the centrosome function.

We thank E. Noguchi (Kyushu University, Fukuoka, Japan) for plasmid pAS404. The English used in this manuscript was revised by K. Miller (Royal English Language Centre, Fukuoka, Japan).

This work was supported by Grants-in-Aid for Specially Promoted Research from the Ministry of Education, Science, and Culture of Japan to T. Nishimoto (08102008) and by the Human Frontier Science Program.

γ-TuRC

γ-tubulin ring complex

ORF

open reading frame; PCM, pericentriolar material

RACE

rapid amplification of cDNA ends

SPB

spindle pole body

ts

temperature sensitive