Portal de Periódicos da CAPES

Molecular analysis of kinetochore architecture in fission yeast

2005; Springer Nature; Volume: 24; Issue: 16 Linguagem: Inglês

10.1038/sj.emboj.7600762

1460-2075

Xingkun Liu, Ian X. McLeod, Scott Anderson, John R. Yates, Xiangwei He,

Fungal and yeast genetics research

Article4 August 2005free access Molecular analysis of kinetochore architecture in fission yeast Xingkun Liu Xingkun Liu Department of Molecular and Human Genetics, Baylor College of Medicine, Houston, TX, USA Search for more papers by this author Ian McLeod Ian McLeod Department of Cell Biology, The Scripps Research Institute, La Jolla, CA, USA Search for more papers by this author Scott Anderson Scott Anderson Department of Cell Biology, The Scripps Research Institute, La Jolla, CA, USA Search for more papers by this author John R Yates III John R Yates III Department of Cell Biology, The Scripps Research Institute, La Jolla, CA, USA Search for more papers by this author Xiangwei He Corresponding Author Xiangwei He Department of Molecular and Human Genetics, Baylor College of Medicine, Houston, TX, USA Search for more papers by this author Xingkun Liu Xingkun Liu Department of Molecular and Human Genetics, Baylor College of Medicine, Houston, TX, USA Search for more papers by this author Ian McLeod Ian McLeod Department of Cell Biology, The Scripps Research Institute, La Jolla, CA, USA Search for more papers by this author Scott Anderson Scott Anderson Department of Cell Biology, The Scripps Research Institute, La Jolla, CA, USA Search for more papers by this author John R Yates III John R Yates III Department of Cell Biology, The Scripps Research Institute, La Jolla, CA, USA Search for more papers by this author Xiangwei He Corresponding Author Xiangwei He Department of Molecular and Human Genetics, Baylor College of Medicine, Houston, TX, USA Search for more papers by this author Author Information Xingkun Liu1, Ian McLeod2, Scott Anderson2, John R Yates2 and Xiangwei He 1 1Department of Molecular and Human Genetics, Baylor College of Medicine, Houston, TX, USA 2Department of Cell Biology, The Scripps Research Institute, La Jolla, CA, USA *Corresponding author. Department of Molecular and Human Genetics, Baylor College of Medicine, Houston, TX 77030, USA. Tel.: +1 713 798 2093; Fax: +1 713 798 8142; E-mail: [email protected] The EMBO Journal (2005)24:2919-2930https://doi.org/10.1038/sj.emboj.7600762 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info Kinetochore composition and structure are critical for understanding how kinetochores of different types perform similar functions in chromosome segregation. We used affinity purification to investigate the kinetochore composition and assembly in Schizosaccharomyces pombe. We identified a conserved DASH complex that functions to ensure precise chromosome segregation. Unlike DASH in budding yeast that is localized onto kinetochores throughout the cell cycle, SpDASH is localized onto kinetochores only in mitosis. We also identified two independent groups of kinetochore components, one of which, the Sim4 complex, contains several novel Fta proteins in addition to known kinetochore components. DASH is likely to be associated with the Sim4 complex via Dad1 protein. The other group, Ndc80–MIND–Spc7 complex, contains the conserved Ndc80 and MIND complexes and Spc7 protein. We propose that fission yeast kinetochore is comprised of at least two major structural motifs that are biochemically separable. Our results suggest a high degree of conservation between the kinetochores of budding yeast and fission yeast even though many individual protein subunits do not have a high degree of sequence similarity. Introduction Central to accurate chromosome segregation is the kinetochore, a multi-protein organelle assembled on centromeric DNA that mediates the interaction between chromosomes and spindle microtubules (MTs). Kinetochores perform multiple functions during chromosome segregation: providing the biochemical affinity between chromosomes and the spindle MTs, generating the physical forces that drive chromosome movement, and acting as sensors of proper MT attachment that couples with the spindle checkpoint pathway. These important biological functions are performed by essentially all types of kinetochores, although in different organisms they have dramatic differences in their composition and morphology—a longstanding enigma in the field of mitosis. The most significant difference among kinetochores is the centromeres, the size of which varies from 125 bp in budding yeast to several megabases in metazoans. In recent studies, major progress in identifying the protein composition of some kinetochores has been made, particularly in budding yeast in which more than 60 distinct proteins have been identified so far (McAinsh et al, 2003). In other species, knowledge of kinetochore proteins is now accumulating at an impressive rate as well (Cleveland et al, 2003). Some kinetochore proteins, such as Ndc80p, are highly conserved in their primary sequences and cellular functions (Zheng et al, 1999). On the other hand, the similarity between other essential kinetochore proteins is less or not at all obvious (McAinsh et al, 2003). Efforts to decipher the architecture of kinetochores have also been initiated, most extensively in budding yeast (De Wulf et al, 2003; Westermann et al, 2003). Many of the kinetochore proteins in budding yeast are organized into discrete complexes which are likely to be structural and functional modules. Studies on the interdependency of kinetochore localization among the components have led to an emerging hierarchy of kinetochore assembly—inner kinetochore proteins that are in contact with DNA, such as CBF3, form the foundation for the mid-layer proteins, represented by three independent complexes, Ndc80, MIND and COMA, and, further, the outer kinetochore components, such as the motor proteins and the DASH complex (McAinsh et al, 2003). Kinetochores from different organisms behave remarkably differently. In budding yeast, kinetochores are attached to spindle MTs throughout almost the entire cell cycle. Therefore, it can be implied that functional kinetochores are always needed. In vertebrate cells, on the other hand, assembly of kinetochores starts in S phase and continues until early mitosis (Amor et al, 2004; Maiato et al, 2004). Thus, a thorough investigation in multiple organisms is necessary to elucidate the general principles of kinetochore function and to distinguish the structural and mechanistic features that are universal to all kinetochores from those that are species-specific. The fission yeast, Schizosaccharomyces pombe, provides an excellent model for the study of kinetochores. With repetitive and complex centromeric DNA spanning from 35 to 100 kilobases and the capacity for binding multiple MTs per kinetochore, fission yeast cells have the complex kinetochores, representative of those found in most eukaryotes. Previous studies have led to extensive knowledge of the sequence and organization of centromeric DNA, as well as the identification of a number of kinetochore-specific proteins in fission yeast (Pidoux and Allshire, 2004). Several intriguing features of kinetochore behavior in fission yeast are noteworthy. The centromeres cluster tightly in the vicinity of the spindle pole body (SPB) throughout the interphase (Uzawa and Yanagida, 1992), so that kinetochores are visualized by fluorescence microscopy as a single dot adjacent to or overlapping the SPB (Goshima et al, 1999; Nabetani et al, 2001; Jin et al, 2002; Pidoux et al, 2003). Electron microscopy (EM) studies have suggested that the tethering of kinetochores to the SPB may be direct, since no MTs were found between the kinetochores and the β-tubulin of SPB (Ding et al, 1997; Kniola et al, 2001). Upon entry into mitosis, a short bipolar spindle is assembled, and sister kinetochores become attached to the dynamic plus ends of spindle MTs and establish a bi-orientation configuration in which the sisters are linked to the opposite SPBs via kinetochore MTs (kMTs). At this stage of mitosis, kinetochores are seen as several discrete dots aligned between the two spindle poles (Uzawa and Yanagida, 1992; Nabetani et al, 2001). The fact that kinetochores are tethered to the SPB implies that functional kinetochores of a certain form must be assembled in interphase. On the other hand, to accommodate their release from SPBs and subsequent attachment to MTs upon entry into mitosis, kinetochores may be biochemically modified. Consistently, the composition of kinetochores changes from interphase to mitosis; some proteins here referred to as constitutive components are localized to kinetochores throughout the cell cycle, while others, here referred to as transient components, are localized to kinetochores only in mitosis. Typical constitutive components include Nuf2, Mis6, Mis12, Sim4, Mal2, and Cnp1 (Goshima et al, 1999; Takahashi et al, 2000; Nabetani et al, 2001; Jin et al, 2002; Pidoux et al, 2003). Typical transient components include Dis1, Alp14, Klp2, Klp5, and Klp6, all of which interact with MTs (Nakaseko et al, 2001; Troxell et al, 2001; Garcia et al, 2002a, 2002b; West et al, 2002). In this study, using affinity purification and mass spectrometry, we identified three kinetochore protein complexes in fission yeast and novel kinetochore components, as well as known ones, as the subunits of these complexes. Two complexes, Sim4 and Ndc80–MIND–Spc7 (NMS), are constitutive components. The third complex, DASH, is transiently associated with kinetochores only in mitosis and is required for precise chromosome segregation. We demonstrated that the Sim4 complex functions as a loading dock for the DASH complex. Results Identification of the DASH complex as a transient kinetochore component The DASH (Dam1, DDD) complex, a 10-subunit complex in the budding yeast, functions in securing the sister kinetochore bi-orientation (Cheeseman et al, 2001; Janke et al, 2002; Li et al, 2002). Inactivation of DASH interferes with biorientation, but not with MT capture. Given that DASH mediates kinetochore/MT interaction and that kinetochores interact with MTs only in mitosis in fission yeast, one can speculate that the postulated S. pombe DASH may be kinetochore localized specifically during mitosis. Among the DASH subunits, three have readily identifiable homologs in S. pombe, one of which, SPBC27.02c, is similar to ScAsk1. We sought to determine the intracellular localization of SPBC27.02c by constructing and expressing a GFP fusion protein from the endogenous promoter as the only functional product of SPBC27.02c. No specific GFP signal was detected in interphase cells. Only in mitotic cells, as indicated by separated SPBs, were localized GFP signals detected (Figure 1A). The multi-dot localization pattern in metaphase cells overlaps closely with the kinetochore marker, Ndc80-CFP (Figure 1A, lower panels). Spindle-like localization (Figure 1A, arrow) and dots in addition to the kinetochore (Figure 1A, arrowhead), but not SPB dots (Figure 1A, middle), were also detected. We thus concluded that SPBC27.02c is a transient kinetochore component and named it Ask1 (or SpAsk1 when needed to be distinguished from ScAsk1). Figure 1.Identification and characterization of DASH complex. (A) Live-cell images of wild-type cells with Ask1-GFP and other markers. The mid and lower panels were digitally enhanced to show the localizations other than kinetochores (arrow and arrowhead). (B) Summary of mass-spectrometric analysis of DASH purifications, indicating the percent sequence coverage for each polypeptide with >5%. Nonspecifically copurified proteins are listed in Supplementary Results. (C) Amino-acid sequences of Dad4 and Spc19 deduced from their genomic DNA sequences. The N-terminal sequence of Spc19 in red illustrates the fragment missing from the incomplete ORF in Sanger Center GeneDB. Download figure Download PowerPoint To identify the postulated SpDASH, we performed tandem affinity purification, using Ask1 as the affinity bait, and identified copurified proteins by mass spectrometry. A fusion of ask1 with the DNA-encoding TAP, a well-established tag for tandem affinity, was expressed under its endogenous promoter as the only functional ask1 gene. To reduce the copurification of nonspecific proteins, we performed the purification at high stringency with high salt concentration (see Materials and methods). Among the proteins identified by mass spectrometry (using a 5% sequence coverage cutoff), nine were repeatedly copurified with Ask1-TAP from cells arrested in mitosis either by the mts2-1 mutation (Seeger et al, 1996) or the nda3-311 mutation (Hiraoka et al, 1984) (Figure 1B; Supplementary data). Among the nine copurified proteins, we identified the other two predicted homologs: Hos2 (later called Dad2) and SPAC14C4.16. Each of the other seven proteins, when aligned pairwise with the ScDASH subunits, shows low but significant sequence similarity (Supplementary Results). We thus adopted the budding yeast protein nomenclature for the homologous proteins in fission yeast. SpSpc19 was not listed as a complete ORF in the Sanger Institute GeneDB database due to a missing start codon, possibly because the first exon was skipped in ORF assignment. We have identified the missing N-terminal by sequencing an Spc19 cDNA (Figure 1C). Mass spectrometry verified the authenticity of the missing N-terminal polypeptide (data not shown). SpDad4 (Figure 1C) was not listed in the GeneDB database either, possibly due to its extremely short length. New ORF numbers have now been assigned: SPCC1223.15c for SpSpc19 and SPBC3B9.22c for SpDad4. To test whether Ask1 and the copurified proteins form a stable complex, we performed reciprocal purification in mitotic cell extracts using SpDam1-TAP or SpDad1-TAP. Each of the postulated DASH subunits was copurified in most or all independent experiments (Figure 1B). We thus conclude that DASH is conserved in fission yeast. SpDASH is assembled and localized onto kinetochores specifically in mitosis SpAsk1-GFP localizes on kinetochores and the spindle only during mitosis (see above). Three other tested DASH components, Dam1, Spc34, and Dad2 (data not shown), exhibited an identical localization pattern, whereas Dad1 did not (see below). To explore the underlying mechanism of its dynamics, we assessed the protein levels of the DASH subunits during the cell cycle. This was performed using nda3cs cells carrying Ask1-TAP that were first arrested at the entry of mitosis by shifting to their restrictive temperature and then allowed to re-enter mitosis synchronously. Cells were harvested at different time points, corresponding to different stages of mitosis (Hiraoka et al, 1984). Cell extracts at each time point were prepared in the presence of trichloroacetic acid (TCA) to cause protein precipitation and prevent degradation. By Western blotting, the total Ask1-TAP protein quantity was found to be constant throughout the cell cycle. In addition, during mitosis (0–30 min in Figure 2A), the intensity of the upper band of Ask1-TAP gradually declined, while the lower band gained prominence and eventually became the only band in interphase (60–120 min in Figure 2A). In budding yeast, Ask1 is phosphorylated extensively (Li and Elledge, 2003). To test whether the upper band represents the phosphorylated SpAsk1, Ask1-TAP was first partially purified from nda3cs cells arrested in mitosis by protein A affinity using IgG resins and then treated with protein phosphatase in the absence or presence of phosphatase inhibitors (PI). The upper band was eliminated by phosphatase treatment but present with the addition of PI, demonstrating that SpAsk1 was phosphorylated specifically during mitosis. By the same assay, the protein levels of Dam1-TAP and Dad1-TAP were also found to be constant, although no obvious phosphorylation bands were detected (data not shown). Figure 2.DASH complex dynamics in mitosis. (A) Mitosis-specific modification(s) of Ask1. Left, Western blotting with PAP antibody to detect Ask1-TAP in total cell extracts of nda3 cells incubated at 20°C and then released into 36°C for times as indicated. Time points of 0–30 min correspond to the M phase; 60–120 min correspond to interphase. Right, partially purified Ask1-TAP was either not treated with λ phosphatase (λPPase) or treated with λPPase in the absence or presence of PI. (B) Typical localization of Ask1-GFP relative to chromosomal DNA and to SPBs in live nda3 cells incubated at 18°C for 10 h. Left panels: a cell that has Ask1-GFP signal only near the SPBs; right panels: a cell that has Ask1-GFP near the SPBs and a dot on an unattached chromosome (arrowhead). (C) Typical images of nda3 cells with localization of Ask1-GFP that is associated with only two (left) or with all the three scattered kinetochore dots (right). Download figure Download PowerPoint Purification of DASH complex by Ask1-TAP or Dam1-TAP was only successful from mitotic cell extracts (Figure 1B). Repeated attempts failed to purify Ask1-TAP and its associated proteins from asynchronous cultures, which were predominantly interphase cells. Indeed, Ask1-TAP protein was stable only in mitotic extracts and was quickly degraded in asynchronous extracts beyond the detection level of Western blotting (data not shown). Ask1-TAP degradation was prevented by protein precipitation through TCA immediately after cell lysis (Figure 2A). These results are consistent with the notion that DASH complex, which is stable in solution, may be assembled during mitosis and that the free Ask1, although present in vivo at a constant level, may be prone to nonspecific protein degradation upon cell lysis. To test whether the binding of DASH to the kinetochores is spindle MT-dependent, we investigated DASH protein localization in nda3cs arrested cells. Over 80% of nda3cs cells had Ask1-GFP (Figure 2B, left) or Dam1-GFP signal (data not shown) localized in the vicinity of the SPBs. Investigation of DASH binding to unattached chromosomes was only possible in a few cells with sufficiently scattered chromosomes. In these cells, Ask1-GFP was occasionally seen to be associated with the unattached chromosome (Figure 2B, right). To better quantify the Ask1-GFP association to unattached kinetochores, we examined nda3 cells carrying Ask1-GFP and Ndc80-CFP arrested in mitosis. Among cells that displayed sufficient kinetochore scattering, that is, with ‘triangular’ Ndc80-CFP dots distribution, 95% (38 out of 40) had Ask1-GFP colocalized with two of the kinetochore dots, but not the third (Figure 2C, left). 5% (2 out of 40) had Ask1-GFP tightly co-localized with all three kinetochores (Figure 2C, right). Together, these results indicate that DASH was bound mostly to the kinetochores associated with the SPBs in nda3 cells, while only a small fraction of the unattached kinetochores had DASH binding. SpDASH is not essential for viability, but necessary for accurate chromosome segregation S. pombe cells in which dad2 (initially called hos2) was deleted were viable, but hypersensitive to high osmotic pressure (Nakamichi et al, 2000). We deleted ask1 and dad1 and found that the cells were viable. Thus, none of the tested DASH subunits is essential for viability, in contrast to ScDASH. To test whether the whole SpDASH complex is dispensable for cell viability, we examined the impact of these deletions on the localization of other DASH subunits. In each deletion strain, all other DASH subunits tested, except Dad1-GFP, lost their kinetochore localization (Figure 3A–C). Uniquely, Dad1-GFP was constitutively kinetochore localized. In anaphase, Dad1-GFP was also localized on spindle MTs similar to, albeit less intensively than, other tested DASH subunits (Figure 3B, left panels, arrowheads). In ask1-Δ and dad2-Δ cells, Dad1-GFP remained kinetochore associated, but was no longer spindle localized (Figure 3B, arrows). However, in dad1-Δ cells, none of the tested DASH subunits was kinetochore associated. These results show that SpDad1 alone is sufficient for its own kinetochore association and that other DASH proteins require each other, and possibly the assembly of the whole complex, for kinetochore association. Furthermore, since dad1-Δ cells are viable and that in these cells DASH fails to be kinetochore localized in these cells, we conclude that the whole DASH complex is dispensable for viability. Figure 3.Characterization of DASH-Δ cells. (A) Dam1-GFP is mislocalized in ask1-Δ or dad1-Δ cells. (B) Dad1-GFP retains its kinetochore localization, but not spindle localization, in dad2-Δ cells. Arrowheads: spindle localization in WT; arrows: the absence of spindle localization in dad2-Δ. (C) Summary of interdependency of DASH proteins for kinetochore localization. (D) Kinetochore mis-segregation (top and middle) or lagging (arrowheads in the bottom) in ask1-Δ cells. (E) Quantification of chromosome loss in dad1-Δ cells with chromosome I labeled with GFP. Left, a cell with equal segregation of chromosome I; middle, a cell with nuclear division, but with chromosome I mis-segregated; right, a cell with an undivided nucleus displaced to one end. The numbers indicate the percentage among the septated cells. (F) Images of a time-lapse sequence (in min) of an ask1-Δ cell undergoing anaphase with a lagging chromosome (arrowhead). Download figure Download PowerPoint Deletion of DASH genes often leads to abnormality in mitosis progression and chromosome segregation. In normal anaphase cells, the segregated kinetochores tightly cluster near the SPBs so that kinetochore proteins appear as two single dots overlapping with SPBs (Nabetani et al, 2001). However, in ask1-Δ cells, 86% of anaphase cells (43/50) in an asynchronous population displayed kinetochore lagging and/or mis-segregation (Figure 3D). To quantify the chromosome loss rate, we generated a strain with dad1-Δ and chromosome I tagged with GFP using the LacO tandem array and LacI-GFP (Rabitsch et al, 2003). In cells with a septum, which have finished mitosis, about 86% displayed a normal pattern of equal segregation of chromosome I and the overall nuclear mass (Figure 3E, left). Less than 4% showed mis-segregation of chromosome I but still had nuclear division (Figure 3E, middle), while 10% had the undivided nucleus displaced to one end (Figure 3E, right). The discrepancy between the high percentage of anaphase cells with lagging kinetochore and the relative low percentage of septated cells with chromosome loss suggests that many anaphase cells correct the defect and eventually execute a successful mitosis. To provide direct evidence for this hypothesis, we followed the mitosis progression in dad1-Δ cells by time lapse microscopy. In all six cells that clearly displayed kinetochore lagging during anaphase, the lagging kinetochore (arrowhead in Figure 3F) caught up with the other two in late anaphase, resulting in a normal mitosis. Identification of a network of interacting kinetochore proteins containing Sim4, Mal2, Mis6, Mis15, Mis17, Dad1, and Fta1–7 Sim4 coimmunoprecipitates with Mis6, suggesting that they might be part of a complex (Pidoux et al, 2003). We sought to identify other potential Sim4 interacting proteins by affinity purification. Using the tandem purification strategy described above, Sim4-interacting proteins were copurified with Sim4-TAP from an asynchronous cell extract and identified by mass spectrometry. Among them, Mis6 and two recently identified Mis6 interacting proteins, Mis15 and Mis17 (Hayashi et al, 2004), were found, validating our purification strategy (Figure 4A; Supplementary data). In addition, a kinetochore protein, Mal2 (Jin et al, 2002), and Dad1, a DASH complex subunit (see above), were also identified, suggesting a close interaction between these proteins. Other copurified proteins were novel, most with no noticeable sequence similarity except for uncharacterized ORFs in other lower eukaryotes (Figure 4A; Supplementary data). Figure 4.Characterizations of the Sim4 complex. (A) Summary of mass-spectrometric analysis of Sim4 complex purifications, indicating the percent sequence coverage for each polypeptide >5%. DASH subunits detected in Dad1-TAP purification are listed in (B). (C) Kinetochore localization of Fta proteins. GFP fusion of Fta1 (top) and Fta2 (bottom) localizes to multiple dots that are between the SPBs (left), and overlap with the kinetochores (right) in live cells in metaphase. (D) ChIP of Fta proteins. Fta2, 3, and 4-GFP proteins are associated with the central core (cnt) and the inner repeats (imr) of the centromere I, but not the outer repeats (otr) or the chromosome arm locus (fbp). Download figure Download PowerPoint To test whether copurification of these proteins was specific, we performed the reciprocal purification using Mal2-TAP. Sim4, as well as most of the above proteins were repeatedly copurified with Mal2-TAP. We thus name the novel proteins that copurify with Sim4 and Mal2, Fta1–7 (Sim4 and Mal2 associating proteins) (Figure 4A; Supplementary data). To verify whether the novel Fta proteins are authentic kinetochore components, we examined their localization with GFP fusion, some together with Cut12-CFP, an SPB marker, or Ndc80-CFP, a kinetochore marker. We constructed GFP fusion for Fta1–5 by inserting GFP at the C-terminus of the chromosomal gene. Fta1, 2, 3, and 4 displayed the characteristic kinetochore localization, namely, multiple dots spread between the two SPBs, overlapping the Ndc80-CFP dots in the early mitotic nucleus (Figure 4C). Fta6 and Fta7 were found to be kinetochore localized in a systemic, genomewide characterization of protein localization (M Yoshida, personal communication). We further tested whether Fta proteins were associated with the centromeric DNA by chromatin immunoprecipitation (ChIP) (Jin et al, 2002). Fta1–5 GFP strains were tested using anti-GFP antibody, and Fta2, Fta3, and Fta4 were found to be associated with the central core cnt and the inner repeats imr region of the centromere, but not with the outer repeats otr region or the chromosome arm locus fbp. These results are identical to those for the known kinetochore components, such as Mal2 (Figure 4D) (Jin et al, 2002). Although Fta1-GFP convincingly displayed a kinetochore localization pattern (Figure 4C, top), chromatin IP of Fta1-GFP had a very weak and not always repeatable association with cen DNA (data not shown), perhaps because GFP on the C-terminal of Fta1 within kinetochores was not easily accessible by the antibodies. We were unable to obtain direct evidence of kinetochore association for Fta5. Results of protein localization and chromatin IP by various tagging strategies (at N- or C-terminal; with -GFP or -TAP as the tag) were inconclusive. However, due to its repeated copurification with Sim4 and Mal2, we tentatively include SPAC1F8.06 in the Fta protein collection to reflect its binding to Sim4 and Mal2 in solution. Further investigation is required to verify its identity as a kinetochore protein. Overall, copurification and protein localization and chromatin IP results lead to the conclusion that an interacting network, which we name as the Sim4 complex, encompasses the previously identified kinetochore proteins, Sim4, Mal2, Mis6, Mis15 and Mis17, and Dad1, as well as a group of novel proteins. The Sim4 complex functions as the loading dock for DASH A DASH component, Dad1, repeatedly copurified with Sim4 and Mal2 (Figure 4A). This result prompted us to postulate that Dad1 is a component of both the DASH and Sim4 complexes. To test this idea, we performed TAP purification using Dad1-TAP. From an asynchronous cell extract, most of the components of both the DASH and Sim4 complexes, but not other kinetochore proteins, copurified with Dad1 (Figure 4A–C). These results provide strong evidence that Dad1 mediates the interaction between the DASH and Sim4 complex. We propose that DASH, when assembled in mitosis, is loaded on kinetochores specifically on the Sim4 complex via Dad1. Supporting evidence was also gained in determining the protein localization of Dad1 and the dependency of other DASH proteins on Dad1 for kinetochore localization (Figure 3A–C). Ndc80 complex, MIND complex, and Spc7 are conserved kinetochore components that physically interact with one another Mis12 and Nuf2 are two constitutive kinetochore proteins. Mutations in these two proteins cause chromosome mis-segregation or complete detachment of kinetochores from the spindle (Goshima et al, 1999, 2003; Nabetani et al, 2001; Obuse et al, 2004). In budding yeast, Mtw1, the Mis12 homolog, and Nuf2 are integral components of the MIND and Ndc80 complexes, respectively (Janke et al, 2001; Wigge and Kilmartin, 2001; De Wulf et al, 2003; Nekrasov et al, 2003). These two complexes are biochemically separable and are assembled on the kinetochores independently (De Wulf et al, 2003). To test whether these two complexes are conserved in fission yeast, we performed tandem affinity purification using SpMis12-TAP and SpNuf2-TAP. From an asynchronous cell extract, eight specific proteins were copurified with SpNuf2-TAP (Figure 5A). In addition to SpNuf2, homologs of three other budding yeast Ndc80 complex components, SpNdc80, SpSpc24, and SpSpc25, were identified. This result shows that the Ndc80 complex is conserved in fission yeast. Figure 5.Characterization of the NMS complex. (A) Summary of mass-spectrometric analysis of reciprocal purification of the complex, indicating the percent sequence coverage for each polypeptide >5%. (B) Coelution of Nuf2-TAP and Mis12-TAP in size exclusion chromatography. SDS–PAGE electrophoresis and Western blotting of fractions of a Superose-6 size exclusion column. The peaks of the elusion of the molecular weight markers are labeled at the top. T