The XMAP215 homologue Stu2 at yeast spindle pole bodies regulates microtubule dynamics and anchorage
2003; Springer Nature; Volume: 22; Issue: 18 Linguagem: Inglês
10.1093/emboj/cdg459
ISSN1460-2075
Autores Tópico(s)Plant nutrient uptake and metabolism
ResumoArticle15 September 2003free access The XMAP215 homologue Stu2 at yeast spindle pole bodies regulates microtubule dynamics and anchorage Takeo Usui Takeo Usui The Paterson Institute for Cancer Research, Christie Hospital NHS Trust, Wilmslow Road, Manchester, M20 4BX UK Antibiotics Laboratory, RIKEN Institute for Discovery Research, Hirosawa 2-1, Wako-shi, Saitama, 351-0198 Japan Search for more papers by this author Hiromi Maekawa Hiromi Maekawa The Paterson Institute for Cancer Research, Christie Hospital NHS Trust, Wilmslow Road, Manchester, M20 4BX UK Search for more papers by this author Gislene Pereira Gislene Pereira The Paterson Institute for Cancer Research, Christie Hospital NHS Trust, Wilmslow Road, Manchester, M20 4BX UK Search for more papers by this author Elmar Schiebel Corresponding Author Elmar Schiebel The Paterson Institute for Cancer Research, Christie Hospital NHS Trust, Wilmslow Road, Manchester, M20 4BX UK Search for more papers by this author Takeo Usui Takeo Usui The Paterson Institute for Cancer Research, Christie Hospital NHS Trust, Wilmslow Road, Manchester, M20 4BX UK Antibiotics Laboratory, RIKEN Institute for Discovery Research, Hirosawa 2-1, Wako-shi, Saitama, 351-0198 Japan Search for more papers by this author Hiromi Maekawa Hiromi Maekawa The Paterson Institute for Cancer Research, Christie Hospital NHS Trust, Wilmslow Road, Manchester, M20 4BX UK Search for more papers by this author Gislene Pereira Gislene Pereira The Paterson Institute for Cancer Research, Christie Hospital NHS Trust, Wilmslow Road, Manchester, M20 4BX UK Search for more papers by this author Elmar Schiebel Corresponding Author Elmar Schiebel The Paterson Institute for Cancer Research, Christie Hospital NHS Trust, Wilmslow Road, Manchester, M20 4BX UK Search for more papers by this author Author Information Takeo Usui1,2, Hiromi Maekawa1, Gislene Pereira1 and Elmar Schiebel 1 1The Paterson Institute for Cancer Research, Christie Hospital NHS Trust, Wilmslow Road, Manchester, M20 4BX UK 2Antibiotics Laboratory, RIKEN Institute for Discovery Research, Hirosawa 2-1, Wako-shi, Saitama, 351-0198 Japan *Corresponding author. E-mail: [email protected] The EMBO Journal (2003)22:4779-4793https://doi.org/10.1093/emboj/cdg459 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info The yeast protein Stu2 belongs to the XMAP215 family of conserved microtubule-binding proteins which regulate microtubule plus end dynamics. XMAP215-related proteins also bind to centrosomes and spindle pole bodies (SPBs) through proteins like the mammalian transforming acidic coiled coil protein TACC or the yeast Spc72. We show that yeast Spc72 has two distinct domains involved in microtubule organization. The essential 100 N-terminal amino acids of Spc72 interact directly with the γ-tubulin complex, and an adjacent non-essential domain of Spc72 mediates binding to Stu2. Through these domains, Spc72 brings Stu2 and the γ-tubulin complex together into a single complex. Manipulation of Spc72–Stu2 interaction at SPBs compromises the anchorage of astral microtubules at the SPB and surprisingly also influences the dynamics of microtubule plus ends. Permanently tethering Stu2 to SPBs by fusing it to a version of Spc72 that lacks the Stu2-binding site in part complements these defects in a manner which is dependent upon the microtubule-binding domain of Stu2. Thus, the SPB-associated Spc72–Stu2 complex plays a key role in regulating microtubule properties. Introduction Microtubules (MTs) are hollow cylinders that assemble from a heterodimer of α- and β-tubulin. MTs are inherently polar structures, so that the two ends of an MT are different. Only the MT minus ends are attached to MT organizing centres such as the centrosome of vertebrate cells and the spindle pole body (SPB) of yeast. MT plus ends are directed away from the centrosome towards the cell periphery. MTs are highly dynamic structures and the ends switch from states of growing to shrinking (called catastrophe) and from shrinking to growing (rescue). The plus ends of MTs are more dynamic than the minus ends (Howard and Hyman, 2003). When measured in vivo, MTs are more dynamic than MTs assembled in vitro from pure tubulin, due to the presence of regulatory factors. Members of the conserved family of MT-binding proteins named Stu2 in budding yeast, XMAP215 in Xenopus, mini-spindles (Msps) in Drosophila, Dis1 and Alp14 in Schizosaccharomyces pombe, DdCP224 of Dictyostellium and the colonic-hepatic tumour-overexpressed gene (ch-TOG) in mammalian cells represent proteins that regulate MT dynamics (Nabeshima et al., 1995; Wang and Huffaker, 1997; Gräf et al., 2000). XMAP215-like proteins regulate MT function by a number of mechanisms. One place of action is the MT plus end. Work in Xenopus extract has shown that XMAP215 stabilizes MT plus ends by opposing the destabilizing activity of the kinesin XKCM1 (Tournebize et al., 2000). In contrast, studies using stabilized MTs identified XMAP215 as an MT-destabilizing factor (Shirasu-Hiza et al., 2003). XMAP215-like proteins also bind to centrosomes and SPBs independently of MTs. In Drosophila, the centrosomal protein D-TACC, named after the human counterpart of the transforming acidic coiled coil (TACC) protein, mediates binding of Msps to centrosomes (Gergely et al., 2000; Lee et al., 2001). If D-TACC levels are reduced, Msps does not efficiently concentrate at centrosomes and the centrosomal MTs are destabilized (Lee et al., 2001). The interpretation of this result is complicated by the fact that Msps and D-TACC are also present at MT plus ends. The three human TACC proteins target ch-TOG to centrosomes and TACC-3 has a role in stabilizing centrosomal MTs (Gergely et al., 2003). A number of studies suggest that defects in ch-TOG and the three TACC proteins may lead to cancer; however, the function of these proteins in mammalian cells is poorly understood (Raff, 2002). The budding yeast Stu2 is associated with SPBs, kinetochores, nuclear and astral MTs, whereas Stu2 probably fulfils diverse functions (Wang and Huffaker, 1997; Chen et al., 1998; He et al., 2001). It is not surprising, therefore, that total depletion of Stu2 leads to a mixture of phenotypes including fewer and less dynamic astral MTs, cell cycle arrest and a failure to elongate the mitotic spindle (Kosco et al., 2001; Severin et al., 2001). Although recent evidence suggests that Stu2 is a plus end-binding MT destabilizer (Kosco et al., 2001; van Breugel et al., 2003), the function of Stu2 at SPBs is not understood. It is only known that Stu2 interacts with the SPB component Spc72 (Chen et al., 1998). The SPB is a large multi-layered structure that is embedded in the nuclear envelope throughout the cell cycle. The cytoplasmic side of the SPB organizes the astral MTs involved in nuclear positioning, whereas the SPB nuclear side assembles the nuclear MTs that have a role in chromosome segregation. The protein Spc72 is only associated with the cytoplasmic side of the SPB. The C-terminus of Spc72 targets the protein to this side of the SPB through binding to the SPB components Nud1 and Kar1 (Pereira et al., 1999; Gruneberg et al., 2000; Schaerer et al., 2001). Besides Nud1, Kar1 and Stu2, Spc72 also interact with the yeast γ-tubulin complex, named the Tub4 complex. Spc72 recruits the Tub4 complex to the cytoplasmic face of the SPB, thereby mediating nucleation of astral MTs (Knop and Schiebel, 1998). The little knowledge we have of the function played by XMAP215-like proteins at centrosomes and SPBs is due in large part to the lack of mutants that specifically affect the centrosome/SPB-associated function while leaving the other functions unaffected. Here, we show that the Tub4 complex and Stu2 bind to adjacent but distinct domains in Spc72. Stu2 and the Tub4 complex are tethered together through their independent associations with Spc72, and they cooperate in organizing astral MTs. Analysis of a SPC72 mutant that fails to bind Stu2 demonstrates that SPB-associated Stu2 is required for astral MTs anchorage and regulation of MT dynamics. Results The 35 most C-terminal amino acids of Stu2 confer binding to Spc72 and astral MT plus ends In order to understand how Stu2 binds to SPBs, we investigated which domain of Stu2 confers binding to the SPB component Spc72. In the yeast two-hybrid system, the 35 most C-terminal amino acids of Stu2 (Stu2853–888) were sufficient to mediate interaction with Spc72 (Figure 1A, lane 3). Consistently, deletion of C-terminal amino acids of Stu2 (Stu21–855) completely abolished interaction with Spc72 (Figure 1A, lane 2). The Stu21–855 two-hybrid construct was functional, since it interacted strongly with STU2648–888 (Figure 1A). The two-hybrid data were confirmed by an in vitro approach. The C-terminal truncated Stu21–855 from insect cells had only a weak interaction with recombinant Escherichia coli-expressed N-Spc72 (Figure 1B, lane 6), whereas Stu2 showed strong interaction (lane 4). Thus, amino acids 853–888 of Stu2 are essential for Spc72 binding. Figure 1.The 35 C-terminal amino acids of Stu2 mediate binding to Spc72. (A) The 35 C-terminal amino acids of Stu2 interact with Spc72 in the yeast two-hybrid system. The domain structure of Stu2 is shown. The MT-binding domain is from amino acids 558–657 (Wang and Huffaker, 1997), the coiled coil domain was predicted from amino acids 697–758 and the low complexity region (LC) from amino acids 766–781. Abbreviations: −, +, ++ symbolize no, good or strong interaction in the two-hybrid system, respectively. (B) Stu21–855 failed to bind to N-Spc72 in vitro. Sf9 cell extracts with Stu2 or Stu21–855 (lanes 1 and 2) were incubated with E.coli-expressed purified GST (lanes 3 and 5) or GST–N-Spc72 (lanes 4 and 6) bound to glutathione–Sepharose beads. After washing, the bound proteins were eluted and analysed by immunoblotting. Download figure Download PowerPoint We then asked whether the Spc72-interaction domain of Stu2 is essential for viability and whether STU21–855 cells are specifically defective in Stu2 binding to SPBs. Using a PCR approach, the chromosomal STU2 was terminated after codon 855. Although STU21–855 cells were viable when grown between 14–37°C (data not shown), cells showed nuclear and astral MT defects (see Supplementary figure 1 available at The EMBO Journal Online). Moreover, localization studies showed that the Stu21–855 protein not only failed to bind to SPBs but also associated with strongly reduced efficiency (79.6% in wild-type and 16.7% in STU21–855 cells) with MT plus ends (see Supplementary figure 2C). Thus, the astral MT defect of STU21–855 cells is likely to arise from the failure of Stu2 to function at both ends of astral MTs. Distinct domains of Spc72 facilitate interaction with the fully assembled Tub4 complex and with Stu2 The requirement for the C-terminal 35 amino acids of Stu2 for its function at both the plus and minus ends of astral MTs probably explains why we failed to obtain STU2 mutants with defects in only one of the two activities. Based on two-hybrid and co-immunoprecipitation experiments, it was suggested that Spc72 interacts with the Tub4 complex and Stu2 (Chen et al., 1998; Knop and Schiebel, 1998; Souès and Adams, 1998). If Stu2 and the Tub4 complex bind to distinct regions of Spc72, it should be possible to construct SPC72 mutants in which the association with Stu2 is defective yet the interaction with the Tub4 complex is unaffected. We therefore mapped the binding sites of Spc72 for the Tub4 complex subunit Spc98 (Geissler et al., 1996) and Stu2. As reported previously (Chen et al., 1998; Knop and Schiebel, 1998), full-length Spc72 interacted with Spc98 and Stu2 in the two-hybrid system (Figure 2A, lane 1). Interaction of Stu2 with Spc72 was relatively weak, which is explained by the recruitment of the Stu2 construct to MTs. Analysis of Spc72 subfragments then showed that the C-terminal Spc72231–622 fragment (Spc72SPB), which interacts with the SPB components Nud1 and Kar1 by two-hybrid (Pereira et al., 1999; Gruneberg et al., 2000), failed to bind to Spc98 and Stu2 (lane 7). This suggests that the N-terminus of Spc72 mediates interaction with Spc98 and Stu2. Indeed, the 267 N-terminal amino acids of Spc72 (N-Spc72) showed two-hybrid interactions with Spc98 and Stu2 (lane 4). N-terminal deletion constructs of Spc72 were then used to address whether Spc98 and Stu2 bind to distinct or overlapping regions of Spc72. Spc72 constructs lacking amino acids 116–230 (Spc72ΔStu2*, line 2) or 176–230 (Spc72ΔStu2, line 3) interacted with Spc98 but not Stu2. Similarly, Spc721–116 (N-Spc72ΔStu2, line 6) bound to Spc98 but not Stu2. In contrast, Spc7299–267 (N-Spc72ΔTub4, line 5) only interacted with Stu2 but not Spc98. These data suggest that the N-terminus of Spc72 carries distinct binding sites for the Tub4 complex and Stu2. Figure 2.The N-terminus of Spc72 has distinct binding sites for Stu2 and the Tub4 complex. (A) Two-hybrid interactions of STU2 and SPC98 with indicated fragments of SPC72. Abbreviations are as in Figure 1A. (B) Anti-HA immunoprecipitations from SPC72 (lanes 1 and 5), SPC72-3HA (lanes 2 and 6), SPC72ΔStu2* (lanes 3 and 7), SPC72ΔStu2*-3HA (lanes 4 and 8), SPC72ΔTub4 (lanes 9 and 11) and SPC72ΔTub4-3HA cells (lanes 10 and 12). For SPC72ΔTub4 and SPC72ΔTub4-3HA cells the rare, slow growing survivors of Figure 3A were used. Cell extracts (lanes 1–4 and 9–10) were incubated with the anti-HA antibodies followed by protein G Sepharose. The washed precipitations (lanes 5–8 and 11–12) were then analysed by immunoblotting with affinity-purified anti-Spc72, anti-Tub4, anti-Spc97 and anti-Spc98 antibodies (lanes 5–8) or only with anti-Spc72 and anti-Tub4 antibodies (lanes 11–12). The Spc72 doublets seen in lanes 8, 10 and 12 are due to partial degradation of Spc72. The asterisks in lanes 1–4 indicate the position of the Spc72 proteins. (C) Yeast total extracts of cells expressing STU2 (−) (pRS426-STU2; pSM564) or STU2-3HA (+) (pRS426-STU2-3HA; pSM669) and chromosomally integrated SPC72-9Myc (lanes 1 and 2) or SPC72ΔStu2-9Myc (lanes 3 and 4) were incubated with anti-HA antibodies. The immunoprecipitates (IP) were probed with anti-HA and anti-Myc antibodies (lanes 5–8). (D) Only the complete Tub4 complex bound strongly to Spc72. Tub4 complex subunits expressed in the indicated combination in Sf9 insect cells were incubated with recombinant and purified GST or GST–N-Spc72 from E.coli. After washing, the eluted proteins were analysed by immunoblotting with the indicated affinity-purified antibodies. (E) In vitro binding of recombinant and purified Tub4 complex and Stu2 (lane 1) from Sf9 cells to E.coli expressed and purified GST (lane 2), GST–N-Spc72 (lane 3), GST–Spc72ΔStu2 (lane 4) and GST–Spc72ΔTub4 (lane 5). Proteins were detected by immunoblotting. Download figure Download PowerPoint Immunoprecipitation experiments were performed to confirm the two-hybrid data. Co-immunoprecipitation of Tub4 complex subunits with Spc72 (Figure 2B, lane 6), Spc72ΔStu2* (lane 8) and Spc72ΔStu2 (data not shown) but not with Spc7299–622 (Spc72ΔTub4, lane 12) was consistent with the notion that the Tub4 complex binding site of Spc72 resided in the 100 N-terminal amino acids and that deletion of the Stu2-binding domain did not affect Tub4 complex binding to Spc72. Moreover, confirming that Spc72ΔStu2 fails to associate with Stu2 in vivo, Stu2-3HA co-immunoprecipitated Spc72-9Myc (Figure 2C, lane 6) but not Spc72ΔStu2-9Myc (lane 8). Although the two-hybrid and immunoprecipitation experiments clearly demonstrated that the Tub4 complex and Stu2 interacted with Spc72, it remained unclear whether these interactions were direct or indirect. In vitro binding experiments were performed to show direct binding of the Tub4 complex and Stu2 to Spc72. Reconstituted Tub4 complex from insect cells (Vinh et al., 2002) interacted strongly with recombinant N-Spc72 (Figure 2D, lane 6). In contrast, Spc72 association with individual subunits (Spc97 and Spc98) and heterodimers (Spc97–Tub4 and Spc98–Tub4) was much weaker (lanes 7-12). Thus, only the fully assembled Tub4 complex binds with high efficiency to Spc72. When N-Spc72 subfragments were tested, Stu2 but not the Tub4 complex interacted with N-Spc72ΔTub4 (Figure 2E, lane 5). In contrast, only the Tub4 complex but not Stu2 bound to N-Spc72ΔStu2 (lane 4). These binding data suggest that the N-terminal 100 amino acids of Spc72 interact directly with the Tub4 complex and amino acids 99–267 of Spc72 with Stu2. The Stu2-binding site is not required for Tub4 complex binding and vice versa. The Tub4 complex but not the Stu2-binding region of Spc72 is essential In our S288c strain background SPC72 is an essential gene (Knop and Schiebel, 1998). This allowed us to address whether the Stu2 or Tub4 complex binding domains of Spc72 were required for viability. Cells from which the chromosomal SPC72 was deleted and which were kept alive by SPC72 on a URA3-based plasmid were transformed with LEU2-based plasmids containing SPC72 derivatives. Cells were then tested for growth on 5-FOA. As the 5-FOA selects against the URA3-based SPC72 plasmid, the LEU2-based SPC72 construct was the only source of SPC72 activity in the 5-FOA resistant colonies. Despite its expression, SPC72ΔTub4 failed to support growth on 5-FOA plates (Figure 3A, sector 3). Only after long incubation did a few, poor-growing colonies expressing SPC72ΔTub4 but not SPC72 appear (Figure 3B, lane 2). Similarly SPC72176–622 cells lacking the Tub4 complex and part of the Stu2-binding regions of Spc72 were unable to grow on 5-FOA (sector 4). Thus, the Tub4 complex binding region of Spc72 is essential for viability. However, the Tub4 complex binding domain by itself was not sufficient to support viability (sectors 7 and 8). Therefore, this domain has to be attached to the C-terminal SPB targeting region of Spc72 (Knop and Schiebel, 1998) in order to fulfil its essential function. Finally, both SPC72ΔStu2* and SPC72ΔStu2 constructs allowed growth on 5-FOA (Figure 3A, sectors 5 and 6, 5-FOA), indicating that the Stu2-binding domain of Spc72 is non-essential. Together, the Tub4 but not the Stu2-binding domain of Spc72 is required for cell viability. Figure 3.Only the Tub4 complex binding domain but not the Stu2-binding site of Spc72 is essential for viability. (A) Δspc72 pRS316-SPC72 cells were transformed with the indicated LEU2-based plasmids. Transformants were grown on YPD and 5-FOA plates for 3 days at 30°C. After 7 days at 30°C, a few 5-FOA resistant SPC72ΔTub4 and SPC72176–622 colonies were obtained (data not shown). (B) 5-FOA resistant colonies of (A) were analysed by immunoblotting for Spc72 expression using anti-C-terminus Spc72 antibodies. Download figure Download PowerPoint Spc72ΔStu2 is specifically defective in Stu2 interaction at SPBs Spc72ΔStu2 was characterized to ensure that it only affected the interaction between Spc72 and Stu2 and not the association of Stu2 with the plus ends of astral MTs. The functional Stu2–4GFP localized in SPC72 wild-type (Figure 4A, arrow head) and SPC72ΔStu2 cells (Figure 4B, arrow) to astral MT plus ends. Analysis of the Stu2–4GFP signal intensity then revealed that Stu2 binding to astral MT plus ends was not altered in SPC72ΔStu2 cells compared with SPC72 wild-type cells (Figure 4C). Thus, the binding of Stu2 to astral MT plus ends is not affected by the SPC72ΔStu2 mutation. Figure 4.Stu2 binds to the plus ends of astral MTs of SPC72ΔStu2 cells. Wild-type (A) and SPC72ΔStu2 cells (B) with STU2-4GFP CFP-TUB1 were analysed for Stu2 localization by fluorescence microscopy. The arrowhead in (A) indicates Stu2 associated with astral MT plus ends. The asterisks in (B) highlight the position of the spindle pole, and the arrows point towards the astral MT plus ends containing Stu2. (C) The relative fluorescence intensity of the Stu2–4GFP signal at astral MT plus ends of cells in (A) and (B) was determined. n > 50. (D) Spc72 is not associated with astral MT plus ends. Using SPC72-4GFP CFP-TUB1 cells, Spc72–4GFP localization was determined by fluorescence microscopy. Scale bars: 5 μm. Download figure Download PowerPoint D-TACC, the functional homologue of Spc72, and Msps interact not only at centrosomes but also at MT plus ends (Lee et al., 2001). A similar interaction of Spc72 and Stu2 at astral MT plus ends would complicate the interpretation of the SPC72ΔStu2 phenotype. Previous localization studies using a SPC72-GFP gene fusion may have missed a weak Spc72 signal at astral MT plus ends (Chen et al., 1998; Knop and Schiebel, 1998). We therefore re-investigated the localization of the fully functional Spc72–4GFP in CFP-TUB1 cells. Under conditions that allowed detection of Stu2–4GFP protein at astral MT plus ends (Figure 4A), it was only detected at SPBs but not along astral MTs or MT plus ends (Figure 4D). Spc72ΔStu2 behaved as Spc72 (data not shown). Thus, Stu2 and Spc72 only co-localize at SPBs. It is likely, therefore, that the SPC72ΔStu2 mutation only affects Stu2 at SPBs but not astral MT plus ends. Astral MTs detach from the SPB of SPC72ΔStu2 cells The phenotype of SPC72ΔStu2 cells was determined to address the function of Stu2 at SPBs. Because astral MT behaviour is cell cycle dependent, we analysed cells in different phases of the cell cycle (Carminati and Stearns, 1997; Vogel et al., 2001). Wild-type SPC72 and SPC72ΔStu2 cells were synchronized in the G1 phase of the cell cycle by α-factor block and release. In G1, SPBs of SPC72ΔStu2 and SPC72 cells had similar numbers of MTs (Figure 5A). A more detailed analysis by time-lapse video microscopy, however, showed that astral MTs detached from the SPB of SPC72ΔStu2 (Figure 5E, arrows) but not SPC72 cells (data not shown). Thus, binding of Stu2 to Spc72 is important for the anchoring of astral MTs to SPBs. Figure 5.The Stu2-binding site of Spc72 fulfils distinct functions during the cell cycle. (A–D) Astral MT number and morphology of synchronized SPC72 wild-type and SPC72ΔStu2 cells with GFP-TUB1 were analysed throughout the cell cycle. n > 100. (A) Astral MTs appear to be normal in SPC7ΔStu2 GFP-TUB1 cells in G1. (B) Pre-anaphase budded SPC72ΔStu2 GFP-TUB1 cells have fewer and detached astral MTs. The arrow highlights a detached astral MT and the arrowhead a short mispositioned spindle without astral MTs. (C and D) SPC72ΔStu2 GFP-TUB1 cells in anaphase. The number of astral MTs associated with the SPB in the bud (C) or in the mother cell body (D) is indicated. The arrows indicate largely extended astral MTs in SPC72ΔStu2 GFP-TUB1 cells. (E) Time-lapse analysis of a SPC72ΔStu2 GFP-TUB1 cell in G1. A time-lapse series of the same cell was taken every 15 s. The arrows highlight a detaching astral MT. (F) Time-lapse analysis of a SPC72ΔStu2 GFP-TUB1 cell in anaphase. A time-lapse series of the same cell was taken every 15 s. The arrows indicate astral MTs that contact the bud cell cortex and then fail to depolymerize. Scale bars: 5 μm. Download figure Download PowerPoint Astral MT defects were also observed in pre-anaphase cells. Twenty-three per cent of pre-anaphase SPC72ΔStu2 cells showed astral MTs that were disconnected from the SPB (Figure 5B, arrow). Time-lapse analysis revealed that these MTs were initially organized from the SPB but subsequently detached (see Supplementary figure 3A). Moreover, 60% of SPC72ΔStu2 cells either lacked (Figure 5B, arrow head) or had fewer astral MTs than SPC72 cells (Figure 5B). These astral MT defects resulted in spindles that were no longer oriented along the mother–bud axis in around 80% of SPC72ΔStu2 cells (Figure 5B, arrow head). SPC72ΔStu2 cells in anaphase did not show detached astral MTs. This either means that the detachment of astral MT is restricted to pre-anaphase cells or that detached astral MTs of anaphase cells are highly unstable. Despite the lack of MT detachment, astral MTs of anaphase SPC72ΔStu2 cells were far from normal. Instead, the SPC72ΔStu2 mutation affected the two astral MT bundles organized of the two SPBs in a different and distinct manner. There was only a slight reduction in the number of astral MTs associated with the SPB, which was located within the bud (Figure 5C). However, in 40–50% of these cells, the MTs were about 10 times longer and curved around the bud tip back into the mother cell (Figure 5C, arrows). Time-lapse video microscopy analysis of these SPC72ΔStu2 cells showed that the astral MTs extended normally into the bud at the beginning of anaphase (Figure 5F, 0–285 s, arrows). Instead of the normal depolymerization that is seen in wild-type cells (Carminati and Stearns, 1997), these bud proximal astral MTs of SPC72ΔStu2 cells did not depolymerize, but slid along the bud cortex as anaphase progressed. In contrast, the mother ward directed SPBs of SPC72ΔStu2 cells were not associated with elongated astral MTs. In fact, these SPBs were more often devoid of detectable astral MTs than the bud-ward directed SPB (Figure 5D; 20% versus 45%). In summary, the Stu2-binding domain of Spc72 fulfils multiple functions during the cell cycle. It is required for the anchorage of astral MTs to SPBs in pre-anaphase cells, for formation of sufficient astral MTs throughout the cell cycle, and for depolymerization of astral MTs in late anaphase. Altered astral MTs dynamics of SPC72ΔStu2 cells The dynamic properties of astral MTs of pre-anaphase SPC72 and SPC72ΔStu2 cells were determined from time-lapse data (see Supplementary figure 3B). Although the astral MTs of pre-anaphase SPC72ΔStu2 cells depolymerized at the same rate as seen in wild-type SPC72 cells (Table I), their growth rate was moderately increased by a factor of 1.3. The most significant change was the 1.5- and 2.3-fold reduction in catastrophe and rescue frequencies, respectively (Table I). These changes resulted in more continuous growth and shrinkage of astral MTs of SPC72ΔStu2 cells (see Supplementary figure 3B). Thus, Stu2 binding to Spc72 at the SPB has a significant and important impact upon the dynamics of astral MTs. Table 1. Astral microtubule dynamics SPC72 (n = 12) SPC72ΔStu2 (n = 12) SPC72ΔStu2 STU21–888–SPC72SPB* (n = 8) a b Rates (μm/min) Growing 1.19 ± 0.59 1.55 ± 0.52 1.23 ± 0.50 P < 0.001 P < 0.001 Shrinking 1.72 ± 0.76 1.76 ± 0.61 1.78 ± 0.70 P = 0.72 P = 0.83 Frequencies (event/s) Catastrophe 0.0147 0.0098 0.0119 Rescue 0.0206 0.0089 0.0150 Event duration (min) Growing 0.74 ± 0.45 0.82 ± 0.46 0.96 ± 0.55 P = 0.43 P = 0.27 Shrinking 0.55 ± 0.25 1.01 ± 0.43 0.77 ± 0.31 P < 0.001 P = 0.017 Pause 0.61 ± 0.16 0.85 ± 0.46 0.67 ± 0.29 P = 0.073 P = 0.24 Event length changes (μm) Growing 0.88 0.96 1.18 Shrinking 1.27 1.74 1.37 Total times (%) Growing 49.6 41.1 52.4 Shrinking 35.3 41.8 35.1 Pausing 15.2 17.1 12.4 a: t-test between SPC72 and SPC72ΔStu2. b: t-test between SPC72ΔStu2 and SPC72 STU21–888-SPC72SPB*. Stu2, the Tub4 complex and Spc72 assemble into a complex and cooperate Both the Tub4 complex and Stu2 are recruited to the cytoplasmic side of the SPB by binding to Spc72 (Chen et al., 1998; Knop and Schiebel, 1998), where they have important roles in astral MT organization (Figures 3 and 5). The close spatial link of the two binding sites in Spc72 (Figure 2) suggests that the Tub4 complex and Stu2 may, in a large complex, cooperate in astral MT organization. To address this possibility, we first asked whether the Tub4 complex and Stu2 are present in common complexes. In such a case, the Tub4 complex subunit Spc97 would be expected to co-immunoprecipitate with Stu2. An anti-HA immunoprecipitate from cells expressing HA-tagged Spc97 contained not only Spc97-3HA but also the Tub4 complex components Spc98 and Tub4, Spc72 and Stu2-3Myc independently of whether logarithmically growing or G1 arrested (α-factor) cells were used (Figure 6A). Thus, the Tub4 complex, Spc72 and Stu2 are part of a larger complex. Figure 6.The Tub4 complex, Stu2 and Spc72 assemble into one complex and act in a cooperative manner. (A) Tub4 complex, Stu2 and Spc72 are part of common complexes. Lysates of logarithmically growing or α-factor arrested SPC97 and SPC97-3HA cells, both with STU2-3Myc, were subjected to anti-HA immunoprecipitation. Precipitated proteins were detected by immunoblotting with the indicated antibodies. (B) Tub4 complex (lanes 1–3 and 7–9) and Stu2-3HA (lanes 4–9), both from Sf9 cells, were incubated with recombinant and purified GST or GST–N-Spc72. After an initial incubation step, anti-HA antibodies were added followed by protein G Sepharose beads. The anti-HA precipitates were analysed by immunoblotting with the indicated antibodies. (C) Genetic interaction between SPC98 and SPC72ΔStu2. Serial dilutions of the indicated yeast cells were grown at the indicated temperatures on YPD plates for 3 days. Download figure Download PowerPoint In vitro binding studies were performed to address how the Tub4 complex, Spc72 and Stu2 interact. Recombinant Tub4 complex and Stu2-3HA were incubated with GST or GST–N-Spc72, followed by the precipitation of Stu2-3HA with anti-HA antibodies. In the presence of GST, the Tub4 complex did not co-precipitate with Stu2-3HA (Figure 6B, lane 8). Addition of N-Spc72 enabled co-immunoprecipitation of the Tub4 complex and Stu2 (Figure 6B, lane 9). Similar results were obtained with N-Spc72-6His (data not shown). This result has two important implicatio
Referência(s)