Journal of Science 111, 2809-2818 (1998) 2809 Printed in Great Britain © The Company of Biologists Limited 1998 JCS3813

SPC72: a spindle pole component required for spindle orientation in the

Sylvie Souès* and Ian R. Adams Laboratory of Molecular Biology, Medical Research Council Centre, Hills Road, Cambridge CB2 2QH, UK *Author for correspondence at present address: Institut de Génétique et de Microbiologie, Bât. 400, Université de Paris-Sud XI, 91 405 Orsay Cédex, France (e-mail: [email protected])

Accepted 7 July; published on WWW 27 August 1998

SUMMARY

The monoclonal antibody 78H6 recognises an 85 kDa capability. Precisely, deletion of SPC72 resulted in component of the yeast spindle pole body. Here we identify a decreased number of astral : early in the cell and characterise this component as Spc72p, the product of cycle only few were detectable, and these were unattached YAL047C. The sequence of SPC72 contains potential coiled- to the spindle pole body in small-budded cells. Later in the coil domains; its overexpression induced formation of large few, if any, remained, and they were unable to polymers that were strictly localised at the outer plaque align the spindle properly. We conclude that Spc72p is not and at the bridge of the spindle pole body. Immunoelectron absolutely required for nucleation per se, but is needed for microscopy confirmed that Spc72p was a component of normal abundance and stability of astral microtubules. these polymers. SPC72 was found to be non-essential for cell growth, but its deletion resulted in abnormal spindle Key words: Yeast, Spindle Pole Body, Astral , positioning, aberrant nuclear migration and defective Microtubule dynamic, Nuclear migration, Mating

INTRODUCTION fails to orient the spindle towards the bud neck, with the consequence that nuclei fail to segregate appropriately (Palmer Microtubules play a key role in and mating of yeast. et al., 1992; Sullivan and Huffaker, 1992). Along with astral Based on morphological considerations, three classes have microtubules, several cytoplasmic motor are been characterised: astral, kinetochore and non-kinetochore implicated in nuclear migration. For instance, dynein heavy (spindle) microtubules. In Saccharomyces cerevisiae, all three chain (encoded by DYN1) and its putative light chain (JNM1) classes originate from a microtubule organising centre known participate in the alignment of the spindle with the mother-bud as the spindle pole body (SPB). The SPB is a multi-layered axis (Li et al., 1993; Eshel et al., 1993; McMillan and Tatchell, disk-like structure which remains embedded in the nuclear 1994; Yeh et al., 1995). Given that dyneins (minus-end directed envelope throughout the cell cycle. The function of the three motors) appear to localise both at the and at main layers is partially understood: a central plaque, which the cell cortex, it has been suggested that dyneins must dock constitutes the SPB core, spans the nuclear envelope, whereas astral microtubules to the bud neck in order to pull the spindle an outer plaque anchors astral microtubules in the ; towards the mother-bud axis. Several of the proteins associated the inner plaque nucleates spindle microtubules in the nucleus with the cell cortex, such as actin, Num1p and the intermediate (Byers and Goetsch, 1975). This capability to nucleate filament-like , Mdm1p, have been shown to sustain the microtubules both at the outer and inner plaques requires the spindle orientation (Palmer et al., 1992; Farkasovsky and association of at least three components: Spc97p and Spc98p Küntzel, 1995; Fisk and Yaffe, 1997). Observation in live cells in the SPB, and Tub4p, the yeast homologue of γ- of DYN1 tagged with green fluorescence protein (GFP) allowed (Sobel and Snyder, 1995; Geissler et al., 1996; Marschall et al., Shaw et al. (1997) to detect highly dynamic astral microtubules 1996; Spang et al., 1996; Knop et al., 1997). Whether or not that were sensing the cell cortex in search of a bud. Thus rather these three proteins are sufficient to achieve nucleation and than on stable attachment, nuclear migration may depend on anchorage of microtubules at the SPB remains to be explored. multiple transient interactions between microtubules and the The role of spindle and astral microtubules has been cell cortex. investigated by genetic studies on TUB2, the unique The half-bridge appears as a horseshoe-like structure encoding β-tubulin in yeast (Huffaker et al., 1988; Sullivan and adjacent to the SPB, and is implicated in SPB duplication Huffaker, 1992; Reijo et al., 1994). Spindle microtubules are during mitosis as well as SPB fusion during mating (Byers involved in segregation (Hoyt and Geiser, 1996). and Goetsch, 1975). Astral microtubules generally nucleate Astral microtubules are essential for proper nuclear migration: at the SPB, but nucleation at the bridge also occurs early in a conditional mutant lacking astral microtubules (tub2-401) the cell cycle. During S phase, following SPB duplication, 2810 S. Souès and I. R. Adams microtubules arising from the bridge are directed towards the with the gene in diploid K842, proper His+ recombinants were bud; they frequently even penetrate into the bud. The function identified by PCR amplification of both ends of the expected insert, of these microtubules is not well established. During mating, using a combination of primers of about 20 bp. after fusion of G1 cells, microtubules originating from the A strain overexpressing Spc72p was prepared by introducing the S. pombe HIS5 selector gene followed by the GAL1-10 promoter bridge connect the SPB of the two partners, and apparently ′ pull the nuclei together before fusion. Several of TUB2 immediately 5 to the SPC72 initiation codon. The strategy used was similar to that described for epitope tagging: the GAL1-10 promoter result in defective astral microtubules, causing improper was amplified from vector pJK404 (Donaldson and Kilmartin, 1996) nuclear positioning and fusion (Huffaker et al., 1988). Loss using primers containing KpnI and XhoI sequences, and inserted into of Kar3p (a minus-end directed motor) results in remarkably the pBS-3xHA-His5 vector at the remaining KpnI and XhoI sites of elongated microtubule bundles, which seem to prevent the polylinker. This Bluescript derivative (pBS-3xHA-His5-Gal1-10) congression of the nuclei (Rose, 1996). In mating cells, was used as template for an amplification reaction. As above, each Kar3p has been localised at both ends of astral microtubules; primer was designed with 18 bp of the template and 45 bp of the it could mediate sliding of astral microtubules as well as their targeted gene (either flanking or including the SPC72 initiation depolymerisation at the SPB (Meluh and Rose, 1990; Endow codon), and proper His+ recombinants were identified by PCR et al., 1994). amplification. The heterozygous strain was cultured at 30¡C in YEP Rout and Kilmartin (1990) prepared a series of monoclonal medium containing 2% raffinose, and Spc72p overexpression was induced by addition of 2% galactose. antibodies directed against an extract enriched in SPB from Deletion of SPC72 open reading frame (ORF) was achieved by a Saccharomyces uvarum. Antibody 78H6 reacted weakly with one-step gene replacement method (Baudin et al., 1993). The an entity of apparent molecular mass 85 kDa, localised at the His3MX6 module was amplified by PCR; again, the forward cytoplasmic side of the SPB (Rout and Kilmartin, 1991). primer (AACACTAATATCAAAAAACTAAGCAAACAACATAA- More recently, we confirmed by matrix-assisted laser GAAAGTTATAGCCTCGACGGATCCCCGGGTTAATTA) and desorption/ionisation (MALDI) mass spectrometry that backward primer (AGAGTGACTGAGTGTTACATTAAATATATTT- extracts enriched in SPB contain a protein of apparent ATATATAAACGTATGATATATCATCGATGAATTCGAGCTCGTT) molecular mass 85 kDa, encoded by YAL047C (Wigge et al., were designed with at least 18 bp of the template and 45 bp flanking + 1998). Here we show that YAL047C encodes the antigen the SPC72 ORF (underlined), and proper His recombinants recognised by 78H6, and we provide evidence about the identified by PCR amplification. After sporulation of the heterozygous strain, tetrads were dissected, and two haploid strains function of this new SPB component. Our results confirm the of opposite mating type (both lacking SPC72) were selected. To localisation of Spc72p at the outer plaque of the SPB and express a Tub1p-GFP fusion protein in the heterozygous strain suggest that it also localises at the half-bridge structure. deleted for SPC72, the TUB1-GFP fusion sequence was recovered Functional characterisation implies that Spc72p either by gap-repair from strain ASF475 (Straight et al., 1997). stabilises or anchors astral microtubules and thus plays a key Centromere-based plasmid pRS314 (TRP/CEN) was digested with role in the complex processes of nuclear migration and KpnI and SacI of the polylinker and used to transform ASF475. Gap- mating. repaired plasmid was recovered from Trp+ yeast and used to transform E. coli TG2 cells by electroporation. One out of 24 E. coli transformants contained a recombinant plasmid with the TUB1-GFP sequence, which was transferred to the polylinker of integrative MATERIALS AND METHODS vector pRS304. Recombinant pRS304 was linearised with Bsu36I in TRP1 and used to transform the heterozygous strain deleted for General methods SPC72. Proper Trp+ recombinants were identified by PCR, and Yeast strains used in this study were all derived from K842 (MATa/α sporulated. Tetrads were dissected and the haploid strain (MATa ade2-1 trp1-1 can1-100 leu2-3,112 his3-11,15 ura3 GAL psi+ ssd1- Spc72∆::HIS5 TRP1::TUB1-GFP) isolated. d2) described in Nasmyth et al. (1990); yeast growth media and yeast techniques used are described in Guthrie and Fink (1991), and pRS Cloning of SPC72 vectors used for DNA manipulation in yeast are described in Sikorski SPC72 was isolated by the method of gap-repair (Orr-Weaver et al., and Hieter (1989). Recombinant DNA methodology in E. coli was 1983). Sequences flanking the SPC72 ORF were amplified to performed as suggested in Sambrook et al. (1989). construct a gap-repair vector: considering the initiation codon as the origin, primers were in the 5′ region at positions −755 (introducing a Gene tagging, overexpression and deletion XhoI site) and at −225 (encompassing the HindIII site). Primers in the Synthetic triple-haemagglutinin epitope YPYDVPDYAGYPYDVPD- 3′ region started at positions 1521 and 2380 (each including an EcoRI YAGSYPYDVPDYAAQCGRSS (3xHA; Tyers and Futcher, 1993) site). Amplifications resulted in 530 and 859 bp fragments that were was used to tag YAL047C or YAL048C by one-step targeted integration sequentially subcloned into pRS316 (URA/CEN) vector. The resulting of PCR-product (Baudin et al., 1993). The 3xHA epitope, followed plasmid was digested with HindIII and Bgl2 (within the EcoRI by a stop codon, and the selective HIS5 marker from fragment) and used to transform K842 cells. Plasmids recovered from Schizosaccharomyces pombe (EUROFAN pFA6a-His3MX6; Wach et Ura+ were used to transform E. coli TG2 cells by al., 1997) were inserted in-frame at the 3′ end of the targeted gene. electroporation. Eight out of 15 E. coli transformants contained a Briefly, starting from a Bluescript vector containing the 3xHA coding recombinant plasmid with the full-length SPC72 ORF. sequence inserted at the NotI site of the polylinker, a stop codon was constructed in-frame with the 3xHA insert by blunting the XbaI site. Mating experiments His3MX6 module was then inserted 3′ to this stop codon, between the To evaluate mating efficiency of strains lacking SPC72, haploids of HindIII and EcoR1 sites of the polylinker. This Bluescript derivative each mating type, either Mata or Matα, and SPC72 or spc72∆ were (pBS-3xHA-His5) was then used as a template for PCR amplification. modified to allow selection of diploids: tryptophan auxotrophy of Each primer was a combination of at least 18 bp homologous with the Matα strains was recovered by integrating in the trp1 genomic locus template and at least 45 bp homologous with sequences flanking the the pRS304 vector linearised in TRP1 with Bsu36I. Likewise, uracyl stop codon of the targeted gene. After homologous recombination auxotrophy of Mata strains was recovered by integrating in the ura3 Spc72p participates in spindle positioning 2811 genomic locus the pRS306 vector linearised in URA3 with Stu1. In a RESULTS typical mating experiment, 106 cells in log phase of each mating type were harvested on a filter that was incubated on YEPD plates at 30¡C Characterisation of SPC72 coding for an 85 kDa for 4 hours. Cells were resuspended in SC-URA-TRP selective component of the SPB medium by sonication, and various cell dilutions were plated onto Screening of a λgt11 expression library of yeast genomic DNA YEPD or SC-URA-TRP plates. Colonies were counted to evaluate with monoclonal antibody 78H6 allowed us to isolate a single mating efficiency. phage. Partial sequencing of the phage insert and a BLAST Morphological observations search identified two potential open reading frames (ORF), The ability of cells to form zygotes was evaluated by light microscopy described in the Saccharomyces Genome Database, namely according to the method of Guthrie and Fink (1991); prior to YAL047C and YAL048C. These ORF are contiguous, on the left observation cells harvested after mating were fixed in methanol:acetic arm of chromosome I, and encode proteins with predicted acid (3:1) for 45 minutes at room temperature, sonicated and washed molecular masses of 72.1 kDa and 75.2 kDa, respectively. twice in phosphate-buffered saline. While western blotting of an enriched SPB preparation Localisations of microtubules, Yal048cp, Spc72p (Yal047cp) and ascribed an apparent molecular mass of 85 kDa to the protein nuclei were determined by immunofluorescence microscopy recognised by 78H6, determination of cellular localisation was according to Kilmartin et al. (1993). For microtubule staining, log- impeded due to the faint immunofluorescence signal. To phase cells were harvested by filtration, immediately washed with 3.7% formaldehyde in 0.1 M potassium phosphate, pH 6.5, and fixed discriminate between the two candidates, YAL047C and in the same buffer for 90 minutes at room temperature. Cells were YAL048C were tagged with 3xHA epitope. Tagging of washed 3 times in 0.1 M potassium phosphate, pH 6.5, once in 1.2 M YAL047C resulted in staining strictly localised at the SPB, sorbitol/0.1 M phospho-citrate, pH 5.9, and the was digested whereas tagging of YAL048C resulted in punctate cytoplasmic with glusulase (10%, v/v) and zymolase (20T, 100 µg/ml) for 90 staining, independent of microtubule location; this eliminated minutes at 30¡C. For SPB component staining (Yal048cp or Spc72p) the 75.2 kDa candidate. This result was confirmed by cells were harvested as above, but washed and fixed for 2 minutes at immunofluorescence analysis of YAL047C overexpression in a room temperature with 3.7% formaldehyde in 1 M sorbitol/0.1 M heterozygous diploid strain: following 78H6 staining there was potassium phosphate, pH 6.5. Cells were washed 3 times in the same a bright signal in induced cells (Fig. 1A). Staining also buffer, once in 1.2 M sorbitol/0.1 M phospho-citrate pH 5.9, and appeared to be strictly localised at the SPB, and was coincident spheroplasting was performed with zymolase (100T, 250 µg/ml) for 1 hour at 30¡C. For concomitant observation of microtubules and with microtubule asters as well as spindle ends. Western Spc72p, the protocol was the same as for SPB component staining, blotting analysis of whole cell extracts confirmed that 78H6 except that fixation was for 20 minutes instead of 2 (in such cases, recognised a protein, overexpressed under galactose induction, SPB component staining was faint, but visible). In all cases, having an apparent molecular mass of 85 kDa (Fig. 1B). spheroplasts were washed in 1.2 M sorbitol/0.1 M phospho-citrate, Recently, we independently confirmed by matrix assisted laser pH 5.9, and a sufficient amount of cells were laid on 15-well slides, desorption/ionisation (MALDI) mass spectrometry that SPB previously coated with polylysine (0.1%). Cells were further fixed at contains a protein of apparent molecular mass 85 kDa encoded −20¡C in methanol for 6 minutes and in acetone for 30 seconds. by YAL047C (Wigge et al., 1998). Taken together, these data Appropriate dilutions of antibody, in phosphate-buffered saline suggested that YAL047C encoded the 85 kDa protein of the containing 1% bovine serum albumin, were added to the spheroplasts SPB that was recognised by monoclonal antibody 78H6; we and incubation was pursued either overnight at 12¡C or for 1 hour at room temperature. Microtubules were labelled with a rabbit anti- formally named it spindle pole component Spc72p. tubulin antibody, Spc72p was labelled with monoclonal antibody 78H6, and tagged proteins with anti-haemagglutinin monoclonal Overexpression of Spc72p-induced formation of antibody 12CA5. Labelling was developed with a secondary antibody: large polymers on the cytoplasmic side of SPB FITC-labelled anti-mouse and/or either Texas-Red- or FITC-labelled An interesting feature of the predicted Spc72p amino acid anti-rabbit (Amersham, Aylesbury, England). Nuclei were sequence was the presence of four regions of potential α- systematically stained by adding 4′,6-diamidino-2-phenylindol helical coiled-coils, which could mediate protein (DAPI, 0.05 µg/ml) directly into the mounting medium. multimerisation (Fig. 2; Berger et al., 1995). Such coiled-coil Cells expressing the TUB1-GFP fusion were observed live by 4 structures are found in Spc110p (Kilmartin et al., 1993), and confocal microscopy (Biorad MRC-600). About 10 exponentially overexpression of Spc110p established its capability to self- growing cells were laid onto a slide previously coated with a thin layer of molten agar, and sealed with a coverslip. assemble, allowing a better understanding of its structural role Cell morphology was also analysed by electron microscopy (EM). in the SPB (Kilmartin and Goh, 1996). To assess whether Cells were fixed and embedded as described by Byers and Goetsch overexpression of Spc72p would induce formation of (1991). Thin sections were cut, stained with 1% uranyl acetate in polymers, cells overexpressing SPC72 were further examined 50% methanol, and treated with Reynolds lead citrate following by EM. After 80 minutes induction, accumulation of electron- standard procedures. Spc72p localisation was also determined by dense material occurred mainly on top of the half-bridge pre-embedding immunoelectron microscopy according to Vandre structure (Fig. 3A), close to the nearest edge of the outer and Burry (1992), with slight modifications (I. R. Adams and J. V. plaque, both in single and paired SPB. At this early stage, no Kilmartin, unpublished). Briefly, cells were fixed in formaldehyde accumulation of material could be seen on the cytoplasmic side for 20 minutes, and spheroplasts were incubated with 78H6 of the outer plaque, but after 120 minutes induction, a larger monoclonal antibody. After washing, monoclonal antibody retained was labelled with Nanogold anti-mouse Fab′ fragment (Nanoprobes, amount of material accumulated, covering both the outer Stonybrooke, NY, USA). Spheroplasts were further fixed with plaque and the half-bridge (Fig. 3B,C). The largest structures glutaraldehyde, Nanogold particles were enlarged by silver resembled balloons (Fig. 3D). A section through these balloons enhancement, samples were treated with osmium, uranyl acetate and revealed a large ring of electron-dense material of regular embedded in resin. shape and even thickness, depicting a highly organised 2812 S. Souès and I. R. Adams A

B

Fig. 1. Monoclonal antibody 78H6 recognises Spc72p, the product of YAL047C. (A) Immunofluorescence microscopy of representative cells: the three upper panels (GAL) represent cells after 2 hours overexpression of YAL047C; the three lower panels (RAFF), cells before induction. Staining was obtained with mAb 78H6, anti-tubulin rabbit polyclonal antibody (α-tubulin) or DAPI. In induced cells, but not in RAFF cells, 78H6 staining colocalises with the microtubule aster, both ends of anaphase spindle and the nucleus rims. Bar, 5 µm. (B) 78H6 antibody detects a protein of apparent 85 kDa only in induced cells, as shown on the immunoblot of whole yeast cell extracts (5-15% gradient polyacrylamide gel electrophoresis in the presence of SDS). The positions of molecular mass markers are shown.

polymer. Serial sections through the whole structure built on astral microtubules to the SPB (Fig. 3B, arrow). To determine top of newly duplicated SPB revealed a round disk of electron whether Spc72p was a component of polymers, we examined dense material associated with the outer plaques (Fig. 3E,H), cells overexpressing Spc72p by immunoelectron microscopy and a ring of dense material connected to the central bridge (Fig. 3I-K): 78H6 antibody decorated both the interior and the between the two SPB (Fig. 3G). In some instances, the top of outside of the balloons, the outer plaque of the SPB and the the outer plaque, free of accumulated material, was enclosed half-bridge. Consistent with these observations, Spc72p is in the balloon, suggesting that polymerisation might have readily detectable at the outer plaque by immunoelectron initiated from the edges of the outer plaque. It was noticeable microscopy either with 78H6 in a yeast extract (Rout and that formation of polymers did not disrupt the attachment of Kilmartin, 1991), or following GFP tagging in whole cells (Wigge et al., 1998). Overall, overexpression of Spc72p led to the formation of ordered polymers attached to the SPB, both 1 at the outer plaque and at the bridge, but distant from the central plaque. 0.8 SPC72 is non-essential, but critical for nuclear 0.6 migration and cell growth 0.4 The growth potential of the heterozygous diploid strain overexpressing Spc72p did not seem affected; it was 0.2 comparable to that of the homozygous SPC72 strain on glucose or galactose medium. To determine whether SPC72 0 was essential for cell viability, the gene was replaced with S. Probability of coiled coil 0 100 200 300 400 500 600 700 pombe His5 in the diploid strain K842 to prepare an heterozygous SPC72/spc72∆ strain. Tetrads obtained after Residue Number sporulation were analysed: all were able to form + Fig. 2. Prediction of coiled-coil structures in Spc72p sequence using colonies, but the two His spores of each tetrad grew poorly the Paircoil computer program (Berger et al., 1995). The graph at 30¡C (Fig. 4). His+ spores actually exhibited ts and cs represents the probability of coiled-coil formation as a function of phenotypes: the rate of growth was decreased dramatically at residue number in the sequence: four regions are of high potential. 37¡C and 23¡C; it was null at 17¡C. Observation through a Spc72p participates in spindle positioning 2813

Fig. 3. Analysis of yeast overexpressing Spc72p by electron microscopy (A-H) and immunoelectron microscopy with 78H6 monoclonal antibody (I-K). In cells induced for 80 minutes, overexpression of Spc72p results in formation of superstructures attached to the SPB bridge (A). In cells induced for 120 minutes, overexpression of Spc72p results in formation of superstructures attached to the SPB outer plaque (E,H) and bridge (D,G). B,C and E-H are two series of thin sections. Overexpression of Spc72p results in formation of superstructures attached to the SPB outer plaque (E,H) and bridge (D,G). The largest structures resemble balloons (D). Spc72p is a component of these superstructures, as attested by decoration with 78H6 antibody (I-K). Astral microtubules remain unperturbed by the superstructures (arrows in B and J). Bars, 0.1 µm. light microscope revealed that the His+ cells were of various SPB (Fig. 5D,E). Taken together, these observations suggest sizes and shapes and presented severe morphological that SPC72 is non-essential, that it is not required for abnormalities (see below Figs 6, 8). While some His+ spores scaffolding of the SPB, but that its absence dramatically resembled small haploid cells having one bud, others were affects the growth capability of cells. larger than diploid cells, and some had multiple buds. The defect in nuclear migration was best seen using Transformation of SPC72-deleted cells with URA/CEN fluorescence microscopy and DAPI staining of an haploid plasmid carrying SPC72 restored growth as well as normal strain lacking SPC72: about a third of the cells appeared to morphological appearance, and counter-selection with 5- have a single nucleus, a third had two nuclei, and the fluoroorotic acid (5-FOA) resulted in the same growth and remaining had up to eleven nuclei (Fig. 6F). With binucleated morphological defects as above. The integrity of the SPB structure in the SPC72-deleted strain was examined by EM. SPC72 deletion did not induce a gross defect of the SPB structure per se. The SPB appeared morphologically normal and its main features could be recognised: outer plaque (Fig. 5E, arrow 1), central plaque spanning the nuclear envelope (Fig. 5E, arrow 2), inner plaque nucleating microtubules (Fig. 5E, arrow 3), and bridge linking the newly duplicated SPB (Fig. 5E, arrow 4). However, no astral microtubules were discernible and most cells contained more than one nucleus, suggesting that nuclear migration might be defective. Perhaps due to the presence of multiple nuclei some rare unusual events were also observed. For instance, a single outer plaque Fig. 4. SPC72 is non-essential but is important for vegetative growth. seemed to be shared by two (Fig. 5A,B), or three (Fig. 5C), Heterozygote diploid strain SPC72/spc72∆ was sporulated at 30¡C, SPB of adjacent nuclei. Possibly due to nuclear fusion events, and tetrads were dissected. The print represents the four spores of each tripolar spindles were also observed which were connecting tetrad dispatched horizontally. All tetrads dissected (about 20) either three SPB, or newly duplicated SPB to two additional contained two slow growing spores that were His+, and lacked SPC72. 2814 S. Souès and I. R. Adams

Fig. 5. Unconventional events in cells lacking SPC72. Cells were grown at 30¡C and analysed by electron microscopy. Most cells contain two or more nuclei. Some cells also exhibit rare events, such as apparent sharing of a single outer plaque by two SPB (A,B, serial thin sections), or even by three SPB (C). In (D and E) (serial thin sections), a single nucleus contains four SPB (including a just-duplicated SPB connected by a bridge); multipolar spindles aberrantly interconnect these SPB. Yet there is no gross defect of the SPB structure: arrows in E point to outer plaque (1), central plaque (2), inner plaque (3) and bridge (4). Bar, 0.1 µm. cells in anaphase, the defect in nuclear migration was nucleus was not visible, and long microtubules were running remarkable (Fig. 6B). The long spindle linking the nuclear along the plasma membrane. Anucleated cells might have lobes was not properly oriented: it was not aligned along the been created by cytokinesis following a defective nuclear bud cell axis, with the consequence that migration of a migration, or by degeneration of very large multinucleated nucleus into the bud was unlikely. Although most of the cells. Typically, spindle formation and nuclear division nuclei remained in the mother cell, a nucleus occasionally seemed to occur normally, but spindles were not oriented entered the bud (Fig. 6D). The mononucleated cells that were properly. Thus, Spc72p appeared to be necessary for proper detected (Fig. 6A) probably originated from such spindle orientation and consequently for effective nuclear stochastic migration of the nuclei. In a few large cells, the migration.

Fig. 6. Nuclear migration defect in the absence of SPC72, analysed by immunofluorescence microscopy. Upper panels represent various cells stained with anti-tubulin antibody (α-tubulin) and lower panels the same cells stained with DAPI. Tubulin staining reveals defects in spindle positioning: most are not along the mother bud axes (B-E). Tubulin staining also reveals that astral microtubules are seldom visible: only in (A) (detached from the SPB) and in (F), early in the cell cycle. In wild-type cells, astral microtubules associate with the SPB staining and extend to the cell cortex throughout the cell cycle (as in cdc13.1 SPC72 cells, Fig. 8A). DAPI indicates that a number of cells contain more than two nuclei: in F, up to 11 are visible in the same focal plane. Nuclear migration defect is obvious in B: two nuclei cohabit in the mother cell. Bar, 5 µm. Spc72p participates in spindle positioning 2815

Fig. 7. Observation by confocal microscopy of live cells lacking SPC72. Microtubules are visualised by GFP-tagging of Tub1p. In (A-C), cells lack SPC72. Astral microtubules are observed only in G1/S phase: long and detached from the SPB (A), or as small asters (B). In addition, no astral microtubules are associated with spindle (C). In (D), cells lacking SPC72 have been rescued with a plasmid carrying SPC72: they exhibit the normal pattern of astral microtubules. Prints are projections of a z-series (9 to 18 steps of 0.5 µm). Bar, 5 µm.

Spc72p is required for astral microtubule function locus (Straight et al., 1997) to allow live observation by The above observations suggested that Spc72p controls one of confocal microscopy. Association of astral microtubules to the functions mediated by microtubules. In fact, nuclear spindles could still not be detected (Fig. 7C), and immunofluorescence analysis revealed that astral microtubules association to dot-like stained SPB was scarce (Fig. 7B). were defective in haploid cells lacking SPC72. Specifically, it Among 41 small budded cells, 30 had no visible astral was not possible to associate astral microtubules to short or microtubule, 8 had short microtubules (Fig. 7B), and only 3 long spindles (Fig. 6B-E). Some short microtubules were had long microtubules, yet detached from the bridge of paired nevertheless pointing toward the axis of the spindle, yet they SPB (Fig. 7A). Furthermore, timed observations revealed that overlapped with DAPI staining, suggesting a nuclear nuclear spindles were static, in contrast with the rapid localisation (Fig. 6C,E). In general, astral microtubules were movements that could be seen in wild type: these exhibited absent in unbudded or small-budded cells, even though a few oscillation of the spindle in the bud neck. Full recovery to short asters of microtubules were visible (Fig. 6F). Long normal phenotype was obtained by transformation of haploid microtubules, usually associated to the bridge of newly SPC72-deleted strain with a CEN vector carrying the wild-type duplicated SPB in wild-type cells, were scarce, and although gene. Rescued cells exhibited astral microtubules that were oriented properly (i.e. directed toward a small bud), they adequately associated, either with newly duplicated SPB and neighboured rather than tied the SPB (Fig. 6A). However, elongating in the bud, or with short spindles and extending both fixation procedures used for immunofluorescence and EM are in the mother cell and the bud neck (Fig. 7D). quite drastic and might have disrupted astral microtubules, To further investigate in the absence raising the possibility that the apparent absence of astral of SPC72, a double mutant was prepared with the temperature- microtubule resulted from an artefact of fixation. Therefore, to sensitive cdc13.1 (Byers and Goetsch, 1974). The evaluate the integrity of microtubules in the SPC72 deleted terminal phenotype of cdc13.1 cells is a short spindle with strain, we integrated the GFP-tagged TUB1 gene at the TRP1 elongated astral microtubules. Immunofluorescence analysis

Fig. 8. Microtubules in cells under cell cycle arrest: observation by immunofluorescence microscopy. Microtubules are stained with an anti- tubulin polyclonal antibody in the top panels (α-tubulin); nuclei are stained with DAPI in the lower panels. Cells in (A) are cdc13.1 that normally express Spc72p; cells in other panels are cdc13.1, lacking SPC72. Observation took place after cell cycle arrest (4 hours shift at 37¡C). In A, astral microtubules are extending into the bud and the mother cell. Without SPC72, astral microtubules are not detectable (B) or seldom and not properly positioned (C-E). In A, the nucleus is close to the bud neck; without SPC72, the nucleus is randomly positioned in the mother cell (B-E). Bar, 5 µm. 2816 S. Souès and I. R. Adams was performed after a 4 hours shift at 37¡C. The cdc13.1 Table 1. SPC72-deleted cells are defective in mating SPC72 cells were arrested with a large bud, a single nucleus, Zygote/total Diploid/total Zygote/ a short spindle close to the bud neck, and astral microtubules Strain Tester (%±s.d.) (%±s.d.) diploid elongating toward the cell edges both in the bud and the mother SPC72 MATa SPC72 MATα 6.3±1.5 4.7±0.6 1.3 cell (Fig. 8A). With the double mutant, the nucleus was spc72∆ MATα SPC72 MATa 5.4±0.8 1.6±0.2 3.4 randomly positioned in the mother cell, the spindle was spc72∆ MATa SPC72 MATα 2.2±0.3 1.0±0.5 2.2 incorrectly oriented, and scattered astral microtubules appeared spc72∆ MATa spc72∆ MATα 1.5±0.5 ≤3×10−5 ≥5×104 non-functional (Fig. 8B-E). Among 44 large budded cells, 22 Zygotes were counted after 4 hours mating. had no visible astral microtubules; 14 had short microtubules Results are expressed as the ratio of budded zygotes observed over total pointing toward the spindle, but concurrent staining with DAPI number of cells counted (means of 4 counts of about 103 cells). suggested a nuclear localisation (Fig. 8B). The remaining cells Number of diploids was estimated after plating on selective medium; had astral microtubule arrays, yet these were apparently results are expressed as ratio of diploid over number of colonies obtained without selection (mean of 3 independent determinations). Reciprocal mating randomly oriented, and unable to accurately position nucleus of SPC72-deficient cells failed to produce diploids despite the number of and spindle, even when pointing toward the bud (Fig. 8C-E). budded zygotes being moderately reduced compared with wild-type testers. Thus, the SPC72-deleted strain had reduced astral microtubules, including at the critical step of nuclear spindle formation. This defect probably prevented proper orientation analysis of a yeast strain overexpressing Spc72p. Overall, this of the spindle toward the bud neck and, consequently, impeded is in full agreement with the preliminary characterisation of legitimate segregation of the nucleus. Undoubtedly, absence of this protein undertaken by Rout and Kilmartin (1991), and our Spc72p drastically affected astral microtubule anchorage recent study by MALDI mass spectrometry (Wigge et al., and/or stability. 1998). Spc72p overexpression resulted in the formation of large polymers both at the bridge and at the outer plaque. SPC72 is necessary for mating Formation of polymers is consistent with the presence of The above observations suggested that Spc72p might be a coiled-coil domains in the Spc72p sequence. Two other component of the bridge, and that, in its absence, astral components of the SPB (Spc110p and Spc42p) contain coiled- microtubules were defective. It is well established that integrity coil domains in their sequence and with both, overexpression of the bridge and of astral microtubules is essential for mating results in formation of polymers initiating from an SPB (Rose, 1996). Compared with reciprocal mating of wild-type substructure where they are readily detectable in wild-type cells, mating of SPC72-deleted cells with wild-type testers cells (Donaldson and Kilmartin, 1996; Kilmartin and Goh, resulted in a three- to fivefold decrease of the mating efficiency, 1996). Thus, it is likely that Spc72p would normally be present but reciprocal mating of cells lacking SPC72 resulted in a at the bridge as well as at the outer plaque. We cannot rule out decrease of at least 103-fold (Table 1). The defect in mating that the polymers included proteins other than Spc72p. could originate from a failure either in the cell or the nuclear However, given the data from immuno-EM analysis, and the fusion step. After mating, cells were examined by light strong signal obtained by immunofluorescence, it is reasonable microscopy to determine whether budded zygotes had formed. to predict that Spc72p was a main (if not unique) element of Budded zygotes could be detected in all mating combinations, our superstructures. Therefore, the consequences of Spc72p even if the number of budded zygotes was somewhat lower overexpression are most probably the direct outcome of an with SPC72-deficient cells compared with reciprocal mating of excess of Spc72p, and not an indirect outcome such as the wild-type cells (Table 1). Therefore the cell fusion step was not induction of the synthesis of a protein which would constitute drastically affected in cells lacking SPC72, inferring that the polymer. Spc72p was most likely implicated in the nuclear fusion step. Spc72p seems to be a regulating element of the SPB machinery, rather than a structural element required for its scaffolding. Based on cryoelectron-microscopy, the DISCUSSION cytoplasmic side of the SPB comprises at least three definite layers: IL1, IL2 and the outer plaque (Bullit et al., 1997). Spc72p was first detected by Rout and Kilmartin (1991) as an Obviously, some of the elements constituting these layers must 85 kDa component of yeast SPB. In the present manuscript, be structural. For instance, the 67 kDa protein recently we provide data that allow the formal identification of Spc72p identified by Brachat et al. (1998), builds in part the as a coiled-coil protein of the SPB localised at the outer plaque cytoplasmic plaque framework: its deletion causes loss of the and the bridge. We also report observations that allow its outer plaque. Deletion of Spc72p did not affect the general function to be determined: rather than being a structural structure of the SPB and none of the layers on the cytoplasmic protein, Spc72p controls proper migration of the nucleus side appeared grossly deficient. On the other hand, deletion of through microtubule anchorage and/or stabilisation. As a Spc72p was clearly deleterious for proper function of the SPB. probable consequence, Spc72p is strictly required for mating. Thus Spc72p, while not a major structural component of the Our results amply confirm that SPB includes an integral SPB, must control its function in part. component of 85 kDa, containing the epitope recognised by SPC72-deleted strains exhibited a phenotype typical of a monoclonal antibody 78H6, and that the gene encoding this nuclear migration defect. Cells were able to assemble a mitotic component is YAL047C (EMBL code: SC12980). Our study spindle, but were deficient in astral microtubules, and nuclear also indicates that Spc72p is expressed on the cytoplasmic side spindles were not correctly oriented. This phenotype is of the SPB. These conclusions are based on comparable to that obtained with β-tubulin conditional mutants immunofluorescence, western blots, EM, and immuno-EM such as tub2-401 (Huffaker et al., 1988; Palmer et al., 1992; Spc72p participates in spindle positioning 2817

Sullivan and Huffaker, 1992). This allele produces preparation of this manuscript. This work was supported by the E.C. multinucleated cells in which nuclear division is mainly grant E243/330; I. R. A. is a recipient of an MRC research completed in the mother cell: they have virtually no astral studentship. microtubules, and their spindles are improperly positioned. Spc98p, Spc97p and Tub4p, all essentials for nucleation of REFERENCES microtubules, are located on the nuclear as well as the cytoplasmic plaques of the SPB (Geissler et al., 1996; Spang Baudin, A., Ozier, K. O., Denouel, A., Lacroute, F. and Cullin, C. (1993). et al., 1996; Knop et al., 1997). An additional element A simple and efficient method for direct gene deletion in Saccharomyces (Spc110p), which acts as a spacer between the inner and the cerevisiae. Nucleic Acids Res. 21, 3329-3330. central plaques, constitutes a docking site for the microtubule Berger, B., Wilson, D. B., Wolf, E., Tonchev, T., Milla, M. and Kim, P. S. (1995). Predicting coiled coils by use of pairwise residue correlations. Proc. nucleating complex in the nucleus (Kilmartin and Goh, 1996; Nat. Acad. Sci. USA 92, 8259-8263. Knop and Schiebel, 1997). It is likely that, to connect the Brachat, A., Kilmartin, J. V., Wach, A. and Philippsen, P. (1998). microtubule nucleating complex to the SPB, one or more Saccharomyces cerevisiae cells with defective spindle pole body outer proteins play a similar role on the cytoplasmic side of the SPB. plaques accomplish nuclear migration via half-bridge-organized SPC72 was not required for nucleation per se, since there were microtubules. Mol. Biol. Cell 9, 977-991. Bullitt, E., Rout, M. P., Kilmartin, J. V. and Akey, C. W. (1997). The yeast some astral microtubules early in the cell cycle (before SPB spindle pole body is assembled around a central crystal of Spc42p. Cell 89, separation), or under cdc13.1 cell cycle arrest. A possibility 1077-1086. would be that Spc72p strengthens the link of nascent Byers, B. and Goetsch, L. (1974). Duplication of spindle plaques and microtubules to the cytoplasmic side of the SPB, as Spc110p integration of the yeast cell cycle. Cold Spring Harbor Symp. Quant. Biol. hooks the nucleating complex on the nucleus rim. Consistent 38, 123-131. Byers, B. and Goetsch, L. (1975). Behavior of spindles and spindle plaques with this hypothesis, the long astral microtubules, which are in the cell cycle and conjugation of Saccharomyces cerevisiae. J. Bacteriol. characteristic of S-phase cells, appeared detached from the 124, 511-523. SPB in the absence of Spc72p. Byers, B. and Goetsch, L. (1991). Preparation of yeast cells for thin-section Under cdc13.1 cell cycle arrest, rare astral microtubules electron microscopy. Methods Enzymol. 194, 602-608. associated with the spindle were disorganised and/or unable to DeZwaan, T. M., Ellingson, E., Pellman, D. and Roof, D. M. (1997). Kinesin-related KIP3 of Saccharomyces cerevisiae is required for a distinct direct the nucleus to the bud neck. Live observations at timed step in nuclear migration. J. Cell Biol. 138, 1023-1040. intervals also suggested that, in the absence of Spc72p, no rapid Donaldson, A. D. and Kilmartin, J. V. (1996). Spc42p: a phosphorylated spindle movements could occur. The nuclear migration defect component of the S. cerevisiae spindle pole body (SPB) with an essential observed with Spc72p-depleted cells is actually similar to that function during SPB duplication. J. Cell Biol. 132, 887-901. Endow, S. A., Kang, S. J., Satterwhite, L. L., Rose, M. D., Skeen, V. P. and of cells lacking motor proteins, such as Kar3p, Kip3p, Salmon, E. D. (1994). Yeast Kar3 is a minus-end microtubule motor protein Dyn1p/Dhc1p or Jnm1p, in which astral microtubules may that destabilises microtubules preferentially at the minus ends. EMBO J. 13, have lost their motility (Palmer et al., 1992; Eschel et al., 1993; 2708-2713. Li et al., 1993; McMillan and Tatchell, 1994; Yeh et al., 1995; Eshel, D., Urrestarazu, L. A., Vissers, S., Jauniaux, J.-C., Van Vliet- DeZwann et al., 1997; Saunders et al., 1997). Thus, Spc72p Reedijk, J. C. Planta, R. J. and Gibbons, I. R. (1993). Cytoplasmic dynein is required for normal nuclear segregation in yeast. Proc. Nat. Acad. Sci. might stabilise astral microtubules by cooperating with motor USA 90, 11172-11176. protein at the SPB. Farkasovsky, M. and Küntzel, H. (1995). Yeast Num1p associates with the That Spc72p is a protein of the bridge implicated in the mother cell cortex during S/G2 phase and affects microtubular functions. J. stability of astral microtubules would explain our results in Cell Biol. 131, 1003-1014. mating experiments. Typically, after cell fusion, astral Fisk, H. A. and Yaffe, M. P. (1997). Mutational analysis of Mdm1p function in nuclear and mitochondrial inheritance. J. Cell Biol. 138, 485-494. microtubules that nucleate at the bridge pull together the nuclei Geissler, S., Pereira, G., Spang, A., Knop, M., Souès, S., Kilmartin, J. V. of opposite mating type prior to nuclear fusion (Byers and and Schiebel, E. (1996). The spindle pole body component Spc98p interacts Goetsch, 1975). Mating is prevented in the absence of astral with the γ-tubulin-like Tub4p of Saccharomyces cerevisiae at the sites of microtubules, as in tub2-401 cells (Huffaker et al., 1988), or microtubule attachment. EMBO J. 15, 3899-3911. when astral microtubule motion is impeded, as in KAR3-deleted Guthrie, C. and Fink, G. R. (1991). Guide to yeast genetics and molecular biology. Methods Enzymol., vol. 194 (ed. C. Guthrie and G. R. Fink). cells (Saunders et al., 1997; Meluh and Rose, 1990; Endow et Academic Press, New York. al., 1994). Disruption of SPC72 abolished mating, yet cellular Hoyt, M. A. and Geiser, J. R. (1996). Genetic analysis of the mitotic spindle. fusion was observed. Thus, Spc72p seems to be required to Annu. Rev. Genet. 30, 7-33. maintain astral microtubules at the bridge and more generally Huffaker, T. C., Thomas, J. H. and Botstein, D. (1988). Diverse effects of to stabilise them, thus controlling their function in mating. β-tubulin mutations on microtubule formation and function. J. Cell Biol. 106, 1997-2010. Kilmartin, J. V., Dyos, S. L., Kershaw, D. and Finch, J. T. (1993). A spacer The authors gratefully acknowledge Dr J. V. Kilmartin for his protein in the Saccharomyces cerevisiae spindle pole body whose transcript continuous support during this project, carried out in his laboratory, is cell cycle-regulated. J. Cell Biol. 123, 1175-1184. and for making available yeast strains and monoclonal antibodies. We Kilmartin, J. V. and Goh, P.-Y. (1996). Spc110p: assembly properties and also would like to thank Dr A. Straight (Department of Physiology, role in the connection of nuclear microtubules to the yeast spindle pole body. School of Medicine, University of California at San Francisco, USA) EMBO J. 15, 4592-4602. Knop, M., Pereira, G., Geissler, S., Grein, K. and Schiebel, E. (1997). The for the generous gift of the ASF475 yeast strain, and Dr A. Brachat γ (Institut für Angewandte Mikrobiologie, Biozentrum, Universität spindle pole body component Spc97p interacts with the -tubulin of Saccharomyces cerevisiae and functions in microtubule organization and Basel, Switzerland) for pFA6a-His3MX6 and pFA6a-kanMX4 spindle pole body duplication. EMBO J. 16, 1550-1564. plasmids. We gratefully acknowledge Dr H. Pelham and Dr R. A. Knop, M. and Schiebel, E. (1997). Spc98p and Spc97p of the yeast γ-tubulin Arkowitz for their support, helpful discussions, and for critical reading complex mediate binding to the spindle pole body via their interaction with of the manuscript. Finally, we would like to thank Dr B. F. Le Bonniec Spc110p. EMBO J. 16, 6985-6995. (INSERM U. 428, Université Paris V, France) for his help during the Li, Y.-Y., Yeh, E., Hays, T. and Bloom, K. (1993). Disruption of mitotic 2818 S. Souès and I. R. Adams

spindle orientation in a yeast dynein mutant. Proc. Nat. Acad. Sci. USA 90, Shaw, S. L., Yeh, E., Maddox, P., Salmon, E. D. and Bloom, K. (1997). 10096-10100. Astral microtubule dynamics in yeast: a microtubule-based searching Marschall, L. G., Jeng, R. L., Mulholland, J. and Stearns, T. (1996). mechanism for spindle orientation and nuclear migration into the bud. J. Analysis of Tub4p, a yeast γ-tubulin-like protein: implications for Cell Biol. 139, 985-994 microtubule-organising center function. J. Cell Biol. 134, 443-454. Sikorski, R. S. and Hieter, P. (1989). A system of shuttle vectors and yeast McMillan, J. N. and Tatchell, K. (1994). The JNM1 gene in the yeast host strains designed for efficient manipulations of DNA in Saccharomyces Saccharomyces cerevisiae is required for nuclear migration and spindle cerevisiae. Genetics 122, 19-27. orientation during the mitotic cell cycle. J. Cell Biol. 125, 143-158. Sobel, S. G. and Snyder, M. (1995). A high divergent γ-tubulin gene is Meluh, P. B. and Rose, M. D. (1990). KAR3, a kinesin-related gene required essential for cell growth and proper microtubule organisation in for yeast nuclear fusion. Cell 60, 1029-1041. Saccharomyces cerevisiae. J. Cell Biol. 131, 1775-1788. Nasmyth, K. A., Adolf, G., Lydall, D. and Seddon, A. (1990). The Spang, A., Geissler, S., Grein, K. and Schieble, E. (1996). γ-tubulin-like identification of a second cell cycle control on the HO promoter in yeast: Tub4p of Saccharomyces cerevisiae is associated with the spindle pole body cell cycle regulation of SW15 nuclear entry. Cell 62, 631-641. substructures that organize microtubules and is required for mitotic spindle Orr-Weaver, T. L., Szostak, J. W. and Rothstein, R. J. (1983). Genetic formation. J. Cell Biol. 134, 429-441. applications of yeast transformation with linear and gapped plasmids. Straight, A. F., Marshall, W. F., Sedat, J. W. and Murray, A. W. (1997). Methods Enzymol. 101, 228-245. Mitosis in living budding yeast: anaphase A but no metaphase plate. Science Palmer, R. E., Sullivan, D. S., Huffaker and T., Koshland, D. (1992). 277, 574-578. Role of astral microtubules and actin in spindle orientation and migration Sullivan, D. S. and Huffaker, T. C. (1992). Astral microtubules are not in the budding yeast, Saccharomyces cerevisiae. J. Cell Biol. 119, 583- required for anaphase B in Saccharomyces cerevisiae. J. Cell Biol. 119, 379- 593. 388. Reijo, R. A., Cooper, E. M., Beagle, G. J. and Huffaker, T. C. (1994). Tyers, M. and Futcher, B. (1993). Far1 and Fus3 link the mating pheromone Systematic mutational analysis of the yeast β-tubulin gene. Mol. Biol. Cell signal transduction pathway to three G1-phase Cdc28 kinase complexes. 5, 29-43. Mol. Cell. Biol. 13, 5659-5669. Rose, M. D. (1996). Nuclear fusion in the yeast Saccharomyces cerevisiae. Vandre, D. D. and Burry, R. W. (1992). Immunoelectron microscopic Annu. Rev. Cell Dev. Biol. 12, 663-695. localisation of phosphoproteins associated with the mitotic spindle. J. Rout, M. P. and Kilmartin, J. V. (1990). Components of the yeast spindle Histochem. Cytochem.40, 1837-1874. and spindle pole body. J. Cell Biol. 111, 1913-1927. Wach, A., Brachat, A., Alberti-Segui, C., Rebischung, C. and Philippsen, Rout, M. P. and Kilmartin, J. V. (1991). Yeast spindle pole body components. P. (1997). Heterologous HIS3 marker and GFP reporter modules for PCR- Cold Spring Harbor Symp. Quant. Biol. 56, 687-692. targeting in Saccharomyces cerevisiae. Yeast 13, 1065-1075. Sambrook, J., Fritsch, E. F. and Maniatis, T. (1989). Molecular Cloning: A Wigge, P. A., Jensen, O. N., Holmes, S., Souès, S., Mann, M. and Laboratory Manual, 2nd edn. Cold Spring Harbor Laboratory Press, Cold Kilmartin, J. V. (1998). Analysis of the Saccharomyces spindle pole by Spring Harbor, New York. matrix-assisted laser desorption/ionization (MALDI) mass spectrometry. J. Saunders, W., Hornack D., Lengyel, V. and Deng, C. (1997). The Cell Biol., 141, 967-977. Saccharomyces cerevisiae kinesin-related motor Kar3p acts at preanaphase Yeh, E., Skibbens, R. V., Cheng, J. W., Salmon, E. D. and Bloom, K. (1995). spindle poles to limit the number and length of cytoplasmic microtubules. Spindle dynamics and cell cycle regulation of dynein in the budding yeast, J. Cell Biol. 137, 417-431. Saccharomyces cerevisiae. J. Cell Biol. 130, 687-700.