The Product of the Spindle Formation Gene sadl + Associates with the Fission Yeast Spindle Pole Body and Is Essential for Viability

Spindle formation in fission yeast occurs by the interdigitation of two microtubule arrays extending from duplicated spindle pole bodies which span the nuclear membrane. By screening a bank of temperature-sensitive mutants by anti-tubulin immunofluorescence microscopy, we previously identified the sad1.1 mutation (Hagan, I., and M. Yanagida. 1990. Nature (Lond.). 347:563-566). Here we describe the isolation and characterization of the sad/+ gene. We show that the sadl.1 mutation affected both spindle formation and function. The sadl + gene is a novel essential gene that encodes a protein with a predicted molecular mass of 58 kD. Deletion of the gene was lethal resulting in identical phenotypes to the sadl.1 mutation. Sequence analysis predicted a potential membrane-spanning domain and an acidic amino terminus. Sadl protein migrated as two bands of 82 and 84 kD on SDS-PAGE, considerably slower than its predicted mobility, and was exclusively associated with the spindle pole body (SPB) throughout the mitotic and meiotic cycles. Microtubule integrity was not required for Sadl association with the SPB. Upon the differentiation of the SPB in metaphase of meiosis II, Sadl-staining patterns similarly changed from a dot to a crescent supporting an integral role in SPB function. Moderate overexpression of Sadl led to association with the nuclear periphery. As Sadl was not detected in the cytoplasmic microtubule-organizing centers activated at the end of anaphase or kinetochores, we suggest that Sadl is not a general component of microtubule-interacting structures per se, but is an essential mitotic component that associates with the SPB but is not required for microtubule nucleation. Sadl may play a role in SPB structure, such as maintaining a functional interface with the nuclear membrane or in providing an anchor for the attachment of microtubule motor proteins. T HE accurate duplication and segregation of genetic information from a parent cell into two daughter cells is a fundamental property of biological systems. Most eukaryotic cells accomplish this complex task by temporally separating different events into a discrete number of interdependent steps. Thus DNA synthesis usually occurs early in the cell cycle and is separated from chromosome segregation by a gap period. Considerable progress has been made in understanding the regulation of these events, in particular the transition of the rate limiting step of commitment to mitosis through the activation of a multicomponent kinase called M-phase Promoting Factor (MPF) t (Nurse, 1990). Despite this increased understanding of the controlling network, our knowledge of the underlying basis of the subsequent chromosome condensation and spindle function Please address all correspondence to Dr. Iain Hagan, 2.205 School of Biological Sciences, University of Manchester, Stopford Building, Oxford Road, Manchester, M13 9PT, United Kingdom. Tel.: 44 61 275 5512. Fax: 44 61 275 5082. 1. Abbreviations used in this paper: MPF, M-phase Promoting Factor; MTOC, microtubule-organizing centre; sad, spindle architecture disrupted; SPB, spindle pole body. remains limited. The morphological aspects of spindle behavior have been well established, but the identities of many spindle components and their specific molecular functions in the spindle remain obscure (Rieder and Salmon, 1994). When spindle formation is initiated by the activation of MPF, the interphase cytoplasmic microtubule network undergoes a radical change. This involves severing existing microtubules, the alteration of the dynamics of microtubules radiating from, and the combined activation and splitting of the principle microtubule-organizing center (MTOC-the centrosome in higher eukaryotes) (Ault and Rieder, 1994; Belmont et al., 1990; Buendia et al., 1992; Gotoh et al., 1990; Shiina et al., 1992; Vale, 1991; McNally and Vale, 1993; Verde et al., 1992). These steps of MTOC activation are considered to be MPF dependent (Centonze and Borisy, 1990; Masuda et al., 1992; Buendia et al., 1992; Verde et al., 1990, 1992). Many recent advances in the identification of spindle components has come from genetic analyses of mutants that affect microtubule or spindle function. This has been particularly apparent with the recent identification of many putative mitotic microtubule motors on the basis of combined DNA sequence homology and mutant phenotype (Goldstein, 1993). © The Rockefeller University Press, 0021-9525/95/05/1033/15 $2.00 The Journal of Cell Biology, Volume 129, Number 4, May 1995 1033-1047 1033 on O cber 9, 2017 jcb.rress.org D ow nladed fom We have been using the fission yeast Schizosaccharomyces pombe as a model system in which to use such a genetic approach towards the identification of spindle components (Yanagida, 1989; Samejima et al., 1993). Fission yeast is well suited for this task as its microtubule cytoskeleton is typically eukaryotic with exclusively cytoplasmic micrombules during interphase which depolymerize as a spindle is formed upon mitotic commitment (Hagan and Hyams, 1988). Spindle micrombules assemble from two spindle pole bodies (SPBs) that are probably activated by MPF-dependent phosphorylation (Masuda et al., 1992). Spindle microtubules are polar and a typical spindle forms by interdigitation of two polarized arrays (McCully and Robinow, 1971; Tanaka and Kanbe, 1986; Masuda et al., 1990; Ding et al., 1993). At the end of an extended anaphase B the spindle breaks down, the SPBs are turned off, and interphase microtubules assemble from two newly activated MTOCs at the cell equator (Hagan and Hyams, 1988; Horio et al., 1991). Mitotic chromosome behavior is also representative of that in higher systems as centromere DNA sequences congress to form a metaphase plate before anaphases A and B (Funabiki et al., 1993). After characterizing the o~and B-mbulin genes (for review see Yanagida, 1987), we isolated the fission yeast 3' tubulin gene, gtbl + by virtue of its identity to the Aspergillus nidulans 7 tubulin gene (Oakley and Oakley, 1989) and showed that it is essential for spindle function (Horio et al., 1991). We have also reported the identification of two fission yeast loci which affect mitotic microtubule patterns in an ostensibly similar fashion when mutated (Hirano et al., 1986; Hagan and Yanagida, 1990). One of these genes, cut7 +, encodes a potential mitotic microtubule motor protein that is a component of the nuclear division apparatus required for mitotic microtubule interdigitation (Hagan and Yanagida, 1990, 1992; Hagan, I., K. Tanaka, and M. Yanagida, unpublished observations). Here we describe the characterization of the second spindle formation gene sad/+ (spindle architecture disrupted [previously referred to as ts549 (Hagan and Yanagida, 1990]). The data indicated that the sad/+ gene encodes an essential protein with predicted molecular mass of 58 kD which associated with the SPB. When fission yeast protein extracts were run on SDS-PAGE, the protein migrated as two distinct bands at 82 and 84 kD. The predicted amino acid sequence of Sadl contained three features of note: an amino terminal acidic region, a 19-amino acid hydrophobic stretch, and a potential cdc2 phosphorylation site consensus sequence. If expressed at elevated levels, Sadl protein accumulated at the nuclear periphery as well as the SPB. Deleting the gene or gross overproduction resulted in the same microtubule distribution as the reduction of Sadl protein levels in the sadl.1 mutant, i.e., a mixture of phenotypes as most cells were unable to form a spindle but others assembled nonfunctional ones. We discuss the possibilities that Sad1 may help to anchor the SPB in the nuclear membrane or act as a docking point for motor proteins. Materials and Methods Strains and Cell Culture The following strains were used: h 9°, 972h-, HMI23 (leuL32 h-) (Gutz et al., 1974) nuc2.663 leul.32 h(Hirano et al., 1988), sadl.1 leu.32 h(formerly ts549) sadl.1 leuL32 his2 h + (Hagan and Yanagida, 1990), and leuL 32/leuL 32 ade6.M210/ade6.M216 his2-/his2ura4.dl8/ura4.dl8 h+/h diploid. Cells were cultured in either rich YPD or minimal EMM2 media (Gutz et al., 1974; Moreno et al., 1990) at the temperatures indicated in the text. For mating and meiotic induction, cells were grown in SSL media at 25°C plus nitrogen, washed in SSL, and incubated in SSL media (Egel, 1971) at 25°C for 9-13 h. Synchronous cultures were generated with a JE-5.0 ehtriator rotor (Beckman Instrs., Fullerton, CA). Genetic Manipulations Standard fission yeast genetic (Gutz et al., 1974; Moreno et al., 1990) and molecular (Sambrook et al., 1989) techniques were used throughout. The plasmid psadl.l was isolated by complementation of the sadJ.l temperaturesensitive defect after transformation with a genomic library in a LEU2 based fission yeast vector (Samejima and Yanasida, 1994). Subcloning the insert defined a minimal 2.l-kb BamH1/Psd insert necessary for complementation in plasmid psadl.10 (Fig. 2 a). The BamHl-PstI fragment was inserted into Bluescript to make plasmid pssdl and the DNA sequence of the entire fragment was determined independently in either direction. This indicated that the reading frame extended beyond the carboxy-terminal PstI site. The 250 nucleotides distal to this site were therefore sequenced to determine the 3' sequence of the gene. To generate the fragment for belle deletion two copies of the ura4 + gene were inserted between the 5' HindIlI site (at position 468 in Fig. 2 b) in pssdl and that Yto the gene in the Bluescript polylinker. Subsequently the XhoI-XhoI fragment from plasmid psadl.7 was inserted into the XhoI site of the polylinker in an orientation relative to the sad/+ BamH1-Hindl~ fragment that corresponded to that in genomic sad/+ to make piasmid pdsadl.2. The BamHl-PstI fragment pdsadl.2 was identical to the similar region of the genome with the exception that the region from nucleotide 469 to 2259 in Fig. 2 c was replaced by two copies of the ura4 + gene. This was used to replace one genomic copy of sad/+ in a diploid strain by homologous recombination (as determined by Southern blotting using the region from the sad/+ distal Pstl site to the second BamHl site as a probe). To place the sad/+ gene under the control of the inducible nmt/+ promoter (Maundrell, 1993), the BamH1-SalI fragment of pssdl was inserted into Blue.Script and the presumptive ATG indicated in Fig. 2 c was mutated in vitro by site directed mutngeuesis to an NdeI site in piasmid pssd6 to generate plasmid pssd6* The entire insert was sequenced to verify that no additional changes had taken place. The polylinker XhoI site of pssd6* was converted to a BamH1 site by blunt end ligatiun of a BamHl linker into the cut plasmid. Subsequently, the SalI fragment ofpsadl.7 including the vector 2 t~ and LEU2 sequences were inserted into pssd6* to generate full-length sad/+ gene in plasmid pSI1. pSI1 can suppress the sadLl mutation. The pSI1 large Ndel-BamH1 fragment was inserted into the Ndel-BamHl site of pREPl to generate pSI2. psi2 was cleaved with XhoI, blunted, ligated with BamH1 linkers, cleaved with BamHl, and religated to generate phtsmid pSI3 in which the sad/+ gene that is under the control of the mm promoter contains only 100 nuclentides of downstream sequences before the nmt polyadenylation signal, pSI3 can suppress the temperature sensitivity of sadl.l mutants when unindnced (basal expression from the wild-type nmt promoter is often sufficiently high to suppress mutant function). Generation of Anti-sadl Antibodies To generate full-len~h Sadl protein in bacteria to use as an immunngen, the Ndel-BamI-ll fragment from pSI3 was inserted into the Ndel-BamI-ll sites of pET-3c (Studier et al., 1990). Despite taking all precautions described in Studier et al. (1990) full-length protein was not detected. Therefore fusion proteins containing the regions NH2 and COOH-terminal to the BgUI site were constructed, SadlFP2 and SadiFP3, respectively. To construct SUdlFF2, the 466-nucleotide Ndel-BgllI fragment from pSI3 was inserted into Ndel BamHl-digested pET-3c. SedlFP3 was made by inserting the t.5-kb BglII-BamHl fragment from pSI3 into the Bam/-I1 site of pET-3a so that the sad/+ open reading frame was inserted in frame with the 2"7 gene 10 encoding sequences in the vector. SadlFP3, migrated with the expected molecular mass of 43 kD, while the NH2-terminal one, SadlFF2, migrated some 10-15 kD more slowly than expected from the amino acid sequence prediction alone. SadlFP3 was incorporated into inclusion bodies upon induction. To purify SadlFP3 for use as an immunogen, or for affinity purification of antibodies from sera, inclusion bodies were purified from induced E. Coil strain BL2t bearing psudlFP3 and piysS + (Studier et al., 1990) and subjected to preparative SDS-PAGE. SedlFP3 was subsequently electroeluted from the relevant gel slice in phosphate buffer (100 mM NaPO4 0.1% SDS, pH 7.2) The Journal of Cell Biology, Volume 129, 1995 1034 on O cber 9, 2017 jcb.rress.org D ow nladed fom and the SDS removed by passage through an Ampure DT column (Amersham Corp., Arlington Heights, IL). SadlFP3 was used to generate two rabbit polyclonal sera (Harlow and Lane, 1988) which were affinity purified with nitrocellulose immobilized SadlFP3 and concentrated to make antibody preparations cAP9 and cAPlO. Protein blotting (Harlow and Lane, 1988) using ECL detection systems (Amersham Corp.) and immunoituorescence microscopy with antibodies purified from various bleeds from each animal showed that all antibody preparations recognized the SPB and the doublet at 82 and 84 kD described in the text. However, only antibodies from the second bleed from rabbit 9 (AP9.2 [affinity-purified antibodies from Rabbit 9 bleed 2]) did not give background bands on protein blots and were therefore used for all applications presen~d in this paper. Antibodies were concentrated from the elution buffer by centrifngation in Sartorious Centristart tubes (cut off 20,000) before use in immunofluorescence (concentrated AP9.2-cAPg.2). Preparation of lqssion Yeast Protein Extracts Fission yeast protein extracts were prepared as described in Moreno et al. (1990).