Correspondence: Kaustuv Sanyal, Molecular Mycology Laboratory, Molecular Biology and Genetics Unit, Jawaharlal Nehru Centre for Advanced Scientific Research, Jakkur Post, Bangalore 560 064, India. Tel: +91 80 2208 2878; fax: +91 80 2208 2766; e-mail: firstname.lastname@example.org
High fidelity chromosome segregation is essential for efficient transfer of the genetic material from the mother to daughter cells. The kinetochore (KT), which connects the centromere DNA to the spindle apparatus, plays a pivotal role in this process. In spite of considerable divergence in the centromere DNA sequence, basic architecture of a KT is evolutionarily conserved from yeast to humans. However, the identification of a large number of KT proteins paved the way of understanding conserved and diverged regulatory steps that lead to the formation of a multiprotein KT super-complex on the centromere DNA in different organisms. Because it is a daunting task to summarize the entire spectrum of information in a minireview, we focus here on the recent understanding in the process of KT assembly in three yeasts: Saccharomyces cerevisiae, Schizosaccharomyces pombe and Candida albicans. Studies in these unicellular organisms suggest that although the basic process of KT assembly remains the same, the dependence of a conserved protein for its KT localization may vary in these organisms.
The precise transmission of the genetic information from one generation to the next during the mitotic cell cycle is extremely important for a eukaryotic organism. This process involves faithful duplication of the whole genome during S phase followed by segregation of the duplicated genome with high fidelity during mitosis. The molecular mechanisms that ensure equal distribution of duplicated chromosomes in mitosis require proper assembly of a large multiprotein complex at the centromere (CEN), known as the kinetochore (KT). The primary function of a KT is to attach the chromosome to the dynamic plus ends of spindle microtubules (MTs), a crucial step in segregation of chromosomes. KTs are also associated with the formation of heterochromatin at the centromeric/pericentric regions and maintenance of cohesion between sister chromatids till anaphase onset (Cleveland et al., 2003; Cheeseman & Desai, 2008). Additionally, a KT is involved in the recruitment of the spindle assembly checkpoint machinery that monitors the KT-MT attachment and initiates signals to prevent cell cycle progression if an error persists. Once all the chromosomes are bi-orientated, separation of two sister chromatids marks the onset of anaphase. Any defect in the KT structure can disrupt KT–MT interaction that may result in an unequal distribution of chromosomes leading to aneuploidy.
Cellular events associated with the mitotic cell cycle
In metazoan cells, the nuclear envelope breaks down during mitosis that allows KT–MT interaction to facilitate bi-oriented chromosomes to arrange on a plane known as the metaphase plate (Nasmyth, 2001; Guttinger et al., 2009). In contrast, the nuclear envelope never breaks down in budding yeasts and thus cells undergo closed mitosis without formation of a metaphase plate (Straight et al., 1997; Sazer, 2005; De Souza & Osmani, 2007). Existence of a metaphase plate is unlikely in Schizosaccharomyces pombe and Candida albicans as well. Interestingly, a semi-open mitosis has been reported recently in fission yeast Schizosaccharomyces japonicus (Aoki et al., 2011; Yam et al., 2011). The nuclear envelope breaks down only during anaphase in this organism. The nuclear envelope virtually breaks down by increasing its permeability during both mitosis and meiosis in S. pombe as well (Asakawa et al., 2010, 2011). These studies indicate that an intermediate type of cell division takes place in fission yeasts.
The microtubule organizing centres (MTOCs) remain outside the nuclear envelope throughout the cell cycle in metazoans. This allows KT–MT interaction to take place only during mitosis. A metazoan KT is typically associated with multiple MTs (20–30 in humans; McDonald et al., 1992). MTOCs, known as the spindle pole bodies (SPBs) in yeast, remain attached to the nuclear envelope throughout the cell cycle in Saccharomyces cerevisiae (Byers & Goetsch, 1975). In addition, KTs remain attached to the MTs throughout the cell cycle in this organism. However, a temporary detachment of chromosomes from MTs occurs (for 1–2 min) at the time of CEN DNA replication in S phase in S. cerevisiae (Kitamura et al., 2007). After completion of CEN replication, a KT reassembles and reestablishes its attachment with MTs. Subsequently, sister CENs precociously separate from each other (Goshima & Yanagida, 2000; Jaspersen & Winey, 2004; Sazer, 2005; Kitamura et al., 2007; Tanaka & Tanaka, 2009). This transient detachment between KT and MT may be specific in organisms (such as S. cerevisiae and C. albicans) in which only one MT interacts with a KT (Ding et al., 1993; Joglekar et al., 2008; Thakur & Sanyal, 2011). Interestingly, SPBs in S. pombe remain outside the nuclear envelope during interphase. Following mitotic initiation, the duplicated SPBs penetrate the nuclear membrane (Ding et al., 1997). The KT is associated with 2–3 MTs in fission yeast (Winey et al., 1995). Therefore, budding yeasts, fission yeasts and metazoans exhibit obvious divergence in timing of commencement of KT–MT interaction, the number of MTs associated with a KT and the fate of the nuclear membrane during the cell cycle.
The kinetochore structure
Early microscopy of mitotic chromosomes in human cells revealed that the human KT is a tri-layered structure (Brinkley & Stubblefield, 1966; McEwen et al., 2007): inner and outer layers that are bridged by a middle layer. Proteins that form the inner layer interact directly with the CEN DNA, while the outer layer proteins form the chromosomal attachment sites of the MT plus ends. Proteins in the middle layer act as linkers between the inner and outer KT (Cheeseman & Desai, 2008). However, in unicellular organisms like yeasts, the structure of a KT cannot be ascertained due to the small-sized cells. Immunostaining of a KT protein in these organisms appears as a single conspicuous focus of clustered KTs at nuclear peripheral regions and close to spindle pole bodies (Meluh et al., 1998; Takahashi et al., 2000; Sanyal & Carbon, 2002). All KTs remain clustered together throughout the cell cycle in S. cerevisiae (Anderson et al., 2009; Duan et al., 2010) and C. albicans (Sanyal & Carbon, 2002; Roy et al., 2011; Thakur & Sanyal, 2011; Fig. 1a). Clustered KTs are found in S. pombe as well except at metaphase where multiple foci of KT proteins were observed (Goshima et al., 1999; Tanaka et al., 2009; Jakopec et al., 2012). Although the exact nature of KT architecture in yeasts is uncertain, various genetic and biochemical studies indicate the presence of functional homologs of several KT proteins at distinct layers of a human KT in these yeasts (Table 1). Determination of relative positions of different proteins at the respective KTs by ‘single molecule high-resolution colocalization’ demonstrates that axial localization of proteins at the KT at distinct phases of mitosis in S. cerevisiae (Joglekar et al., 2009) and humans is largely conserved (Wan et al., 2009; Fig. 1b). However, such studies are yet to be carried out in S. pombe and C. albicans. Nevertheless, the difference in the cross-linking time of KT proteins of C. albicans with CEN chromatin indicates a structural similarity between C. albicans (Sanyal et al., 2004; Roy et al., 2011; Thakur & Sanyal, 2011) and metazoans KTs.
Table 1. Kinetochore proteins/protein complexes in yeasts
Dynamics of assembly of KT proteins is dissimilar in yeasts and metazoans. In metazoans, only the CEN-specific histone H3 variant and an inner KT-associated super-complex, commonly known as constitutive centromere-associated network, remain localized at the KT throughout the cell cycle (Foltz et al., 2006; Liu et al., 2006; Okada et al., 2006). Localization/delocalization dynamics of middle and outer KT proteins is specific to stages of the cell cycle. For example, a middle KT protein and a MT interacting protein are loaded at the KT at late interphase and delocalize from the KT during transition of late anaphase to telophase in metazoans (Liu et al., 2006; Cheeseman & Desai, 2008; Cheeseman et al., 2008). In contrast, proteins from all layers of a KT exhibit constitutive localization at the CEN in S. cerevisiae (Meluh et al., 1998; Goshima & Yanagida, 2000) and C. albicans (Sanyal & Carbon, 2002; Roy et al., 2011; Thakur & Sanyal, 2011). All the outer KT proteins of S. pombe localize at the CEN only during mitosis except one component, which remains localized at the KT throughout the cell cycle (Liu et al., 2005; Sanchez-Perez et al., 2005).
The organization of yeast centromeres
Organization of CENs in different fungi including several yeast species can be classified into three categories: point, large regional and small regional CENs (Roy & Sanyal, 2011; Sanyal, 2012). S. cerevisiae has short point CENs (< 400 bp) with conserved DNA motifs for protein binding, and thus, they are genetically defined (Fitzgerald-Hayes et al., 1982; Hieter et al., 1985). In contrast, S. pombe has longer regional CENs (≥ 40 kb) consisting of repetitive as well as unique DNA elements (Clarke et al., 1986; Nakaseko et al., 1987; Fishel et al., 1988; Takahashi et al., 1992; Steiner et al., 1993; Baum et al., 1994; Wood et al., 2002). C. albicans possesses small regional CENs that span a 3- to 5-kb unique DNA sequence without any repeat elements (Sanyal et al., 2004). Unlike point CENs, regional CENs are epigenetically defined as they do not possess any exclusive CEN-specific protein binding sequence motifs (Steiner & Clarke, 1994; Baum et al., 2006).
A series of experimental evidence gathered from (1) in silico analysis, (2) genetic analysis of KT localization interdependence, (3) biochemical purification of protein complexes and (4) advanced microscopic observations facilitate a comparative analysis of the process of KT assembly in S. cerevisiae, S. pombe and C. albicans – each having a distinct class of CENs as discussed above. Several genetic and biochemical studies identified > 60 proteins that are present at the KT in S. cerevisiae. In contrast, fewer studies were performed on the KT proteins in C. albicans and S. pombe. Thus, we mostly restrict this comparative analysis to only a few KT protein families and their known interacting partners that were studied in all three yeasts – the CENP-A, CENP-C, Mis12 and Dam1 complex. We compare and contrast the processes that lead to KT–MT interaction to facilitate chromosome segregation in these organisms.
Centromeric chromatin properties in yeasts
CEN chromatin properties have been studied in different yeasts. In S. cerevisiae, partial micrococcal nuclease (MNase) digestion along with DNase I digestion of chromatin revealed that there are more distinct ladder patterns at CEN chromatin as compared with that in bulk chromatin (Bloom & Carbon, 1982). In this experiment, mapping exact cleavage sites discovered a distinctly protected region of 220–250 bp of CEN chromatin flanked by a highly phased nucleosome structure with several nuclease sensitive sites. On the other hand, S. pombe and C. albicans contain unusual CEN chromatin. Partial MNase digestion yielded canonical approximately 150-bp ladder patterns in bulk chromatin, while smeary patterns were visible when probed with core CEN regions in S. pombe (Polizzi & Clarke, 1991; Song et al., 2008) and C. albicans (Baum et al., 2006). Thus, CEN chromatin properties seem to be different from canonical H3 chromatin. All CENs are marked by a CEN-specific histone H3 variant – CENP-A. CENP-A molecules replace histone H3 molecules either partially or fully at the CENs in all these three yeast species (Meluh et al., 1998; Takahashi et al., 2000; Sanyal et al., 2004; Burrack et al., 2011). The assembled KT proteins at the CEN may also confer protection against MNase (Song et al., 2008). A recent in vitro study suggested that a complex of CENP-S-T-W-X forms a unique structure of CEN chromatin (Nishino et al., 2012). The homologs of these proteins were identified and characterized in different yeasts as well (Schleiffer et al., 2011; Smith et al., 2011; Bock et al., 2012; Fukagawa, 2012). Incorporation of this complex that form noncanonical nucleosomes also may contribute to the unique structure of CEN chromatin.
The inner kinetochore
Although the process of KT assembly has been shown to be species specific, a common feature of the functional CENs is the existence of a CEN-specific histone H3 variant, CENP-A (Meluh et al., 1998; Takahashi et al., 2000; Sanyal & Carbon, 2002). Inner KT assembly is considered to be initiated by CENP-A deposition. CENP-A recruitment can occur through multiple pathways, which involve several genetic and epigenetic factors. Recruitment of CENP-A takes place at different stages of the cell cycle. It occurs during S phase and anaphase in S. cerevisiae (Pearson et al., 2004; Shivaraju et al., 2012), at S and G2 phases in S. pombe (Chen et al., 2003; Takayama et al., 2008) and at least in anaphase in C. albicans (Shivaraju et al., 2012). Further experimentation is required to investigate whether CENP-A deposits at early S phase when the CEN DNA is replicated in C. albicans (Koren et al., 2010). An evolutionarily conserved nonhistone DNA-binding chaperone Scm3/HJURP is an essential component for KT assembly. This family of proteins has the propensity to bind to the A-T rich CEN DNA and contains a histone chaperone domain, which is required for Cse4/H4 deposition in vivo (Xiao et al., 2011). Scm3 is required for CENP-A deposition at the CEN both in S. cerevisiae and S. pombe (Camahort et al., 2007; Mizuguchi et al., 2007; Stoler et al., 2007; Pidoux et al., 2009; Williams et al., 2009). Moreover, over-expression of Scm3 results in a reduction in Cse4 at the CEN in S. cerevisiae (Mishra et al., 2011). Although Scm3 is required for Cse4 localization at the CEN, but its own localization at the CEN is independent of Cse4 in both S. cerevisiae and S. pombe (Williams et al., 2009; Luconi et al., 2011). Similarly, another KT protein essential for CENP-A localization is CENP-C. The localization of CENP-A is dependent on CENP-C in both S. pombe (Tanaka et al., 2009) and C. albicans (Thakur & Sanyal, 2012). In addition to these proteins, epigenetic regulation of CENP-A deposition (reviewed in Roy & Sanyal, 2011) has been demonstrated in S. pombe (Steiner & Clarke, 1994) and C. albicans (Baum et al., 2006).
Ndc10, a part of the point CEN-specific CBF3 complex, has been shown to influence the recruitment of most of the KT proteins including CENP-A in S. cerevisiae (Ortiz et al., 1999; Russell et al., 1999; Goshima & Yanagida, 2000; He et al., 2001; Janke et al., 2001, 2002). It is not clear that Ndc10 is required only in S. cerevisiae because an obvious homolog is not identified in S. pombe or C. albicans. On the other hand, Ams2 at S phase (Chen et al., 2003) and Hip1 at G2 phase (Takayama et al., 2008) influence CENP-A loading in S. pombe. The cell cycle phase–specific loading of CENP-A has also been shown to be affected by Mis6 through Sim3 in S. pombe (Takahashi et al., 2000; Dunleavy et al., 2007). Interestingly, proteins from the middle and outer KT affect the localization of CENP-A in C. albicans (Roy et al., 2011; Thakur & Sanyal, 2012). The Dam1 complex, a fungal-specific outer KT protein complex, which has no known role in CENP-A recruitment in S. cerevisiae or in S. pombe, influences the localization and stability of CENP-A in C. albicans (Thakur & Sanyal, 2012).
Members of the evolutionarily conserved CENP-C family contain a c. 25-amino acid-long conserved region, known as the CENP-C box, which is essential for its KT localization (Meluh & Koshland, 1995; Yu et al., 2000; Suzuki et al., 2004). CENP-C localization at the KT is mediated by CENP-A in both S. cerevisiae (Westermann et al., 2003) and S. pombe (Tanaka et al., 2009). CENP-C requires Mis12 for its recruitment at the KT in both S. cerevisiae (Westermann et al., 2003) and C. albicans (Roy et al., 2011). Ndc10 and Nnf1 influence CENP-C localization in S. cerevisiae (Meluh & Koshland, 1997; Collins et al., 2005). However, the dependence of CENP-C on Nnf1 has not been studied in S. pombe and C. albicans. Interestingly, subunits of the Dam1 complex are essential for CENP-C localization at the KT in C. albicans (Thakur & Sanyal, 2012).
The middle kinetochore
The Ndc80-MIND-Spc105 (NMS) super-complex
The yeast counterpart of the KNL1-Mis12-Ndc80 (KMN) network, identified in higher eukaryotes, consists of the Ndc80 complex, MIND/Mis12 complex and Spc105/Spc7 complex.
The Ndc80 complex
The requirement of CENP-A for KT localization of the Ndc80 complex is similar in budding yeasts, S. cerevisiae (Collins et al., 2005) and C. albicans (Burrack et al., 2011). Moreover, Cnn1/CENP-T and Ndc10 were reported to influence the assembly of the Ndc80 complex in S. cerevisiae (He et al., 2001; Janke et al., 2001; Schleiffer et al., 2011; Bock et al., 2012; Nishino et al., 2012). Middle KT components including Mis12 and Nnf1 were shown to affect the localization of this complex at the KT (Westermann et al., 2003). In S. pombe, dependence as well as localization of the Ndc80 complex is not well established. The Dam1 complex subunits influence the loading of Nuf2, a constituent of the Ndc80 complex in C. albicans (Thakur & Sanyal, 2012).
The MIND/Mis12 complex
CENP-A plays an important role in recruiting Mis12 at the KT both in S. cerevisiae (Pinsky et al., 2003; Westermann et al., 2003; Collins et al., 2005) and C. albicans (Burrack et al., 2011; Roy et al., 2011) but Mis12 and CENP-A are independent of each other for their KT recruitment in S. pombe (Takahashi et al., 2000).
Ndc10 is essential for the KT localization of each of the constituents of the MIND complex in S. cerevisiae (Goshima & Yanagida, 2000; Nekrasov et al., 2003; Pinsky et al., 2003). KT localization of the Mis12 complex is independent of Spc105 in S. cerevisiae (Pagliuca et al., 2009) but Mis12, Mis13/Dsn1 and Mis14/Nsl1 require Spc7 and Sos7 for their KT localization in S. pombe (Kerres et al., 2007; Pagliuca et al., 2009; Jakopec et al., 2012). Depletion of a subunit of the Dam1 complex affects Mis12 localization in C. albicans (Thakur & Sanyal, 2012).
The Spc105/Spc7 complex
The Spc105 complex of S. cerevisiae consists of two subunits, which are Spc105 and Kre28. Ndc10 influences KT recruitment of both the components of this complex (Nekrasov et al., 2003; Pagliuca et al., 2009). The recruitment of Spc105 at the KT is independent of the MIND and Ndc80 complex in S. cerevisiae (Pagliuca et al., 2009).
However, Spc7/Spc105 forms complex with Sos7, which has been identified recently as a KT protein in fission yeast S. pombe (Jakopec et al., 2012). Spc7 and Sos7 are interdependent for their KT localization (Jakopec et al., 2012). Both the proteins are dependent on Mis12 for their loading at the KT (Kerres et al., 2007; Jakopec et al., 2012).
The outer kinetochore
The Dam1 complex
The Dam1 complex is essential for cell viability and localized at the KT throughout cell cycle in both budding yeasts, S. cerevisiae (Hofmann et al., 1998; Cheeseman et al., 2001a, b; Enquist-Newman et al., 2001) and C. albicans (Burrack et al., 2011; Thakur & Sanyal, 2011). CENP-A influences the KT recruitment of this complex in both the budding yeasts (Collins et al., 2005; Burrack et al., 2011).
In contrast to budding yeasts, the Dam1 complex is nonessential for cell viability in fission yeast S. pombe. Moreover, except Dad1, other subunits of this complex localize at the KT only during mitosis in S. pombe (Liu et al., 2005; Sanchez-Perez et al., 2005). The recruitment of the Dam1 complex is affected by Ndc10, Mis12 and Ndc80 in S. cerevisiae (He et al., 2001; Li et al., 2002; Scharfenberger et al., 2003; Collins et al., 2005; Pagliuca et al., 2009), whereas localization of the Dam1 complex is controlled by the Mis6 complex proteins in S. pombe (Liu et al., 2005; Sanchez-Perez et al., 2005).
In this review, we compared the process and sequence of events during KT assembly in three different ascomycetous yeasts, each carrying a specific type of CEN. While similarities and differences in KT assembly in these organisms are evident, some key questions need to be experimentally addressed.
Ndc10 is the key determinant in KT assembly in S. cerevisiae. Is there a functional homolog of Ndc10 in organisms (such as C. albicans and S. pombe) possessing sequence-independent regional CENs? The requirement of Scm3 for loading of CENP-A is found to be similar in S. cerevisiae and S. pombe but not yet studied in C. albicans. The localization dependence between Ndc80 and CENP-A has been examined in S. cerevisiae and C. albicans but not in S. pombe. The roles of an inner KT protein Mis6/Ctf3 and a middle KT protein Spc105/Spc7 in KT assembly have been studied in S. cerevisiae and S. pombe. The identification and characterization of the functional homologs of these proteins in C. albicans will improve our knowledge of KT assembly in this yeast.
The requirement of the Dam1 complex for assembly of a KT also differs between two budding yeasts, S. cerevisiae and C. albicans. The Dam1 complex requires components of inner and middle KT for its KT localization in S. cerevisiae but not vice versa. In contrast, depletion of the Dam1 complex results in the disruption of KT architecture and destabilization of CENP-A in C. albicans (Thakur & Sanyal, 2012). What will be the consequence of the Dam1 complex depletion on KT architecture and stability of KT proteins in S. cerevisiae and S. pombe? In S. cerevisiae, Dam1 can form MT attachment site if it is targeted by tethering to an ectopic noncentromeric DNA sequence (Kiermaier et al., 2009; Lacefield et al., 2009). It will also be interesting to study what happens if Dam1 is targeted to such an ectopic location in S. pombe or C. albicans where the CEN formation is epigenetically regulated.
Points to ponder
It is important to note that the localization dependence studies were not performed uniformly as the sensitivity of quantitative measurement techniques improved significantly over the years. Moreover, the methods used to assay KT localization dependence are sometimes not mentioned clearly, and in many occasions, the methods are rather qualitative than quantitative. For example, the CENP-A independent localization of Mis12 at the CEN in fission yeast has been claimed based on an experiment that was not shown (Takahashi et al., 2000). Unfortunately, this information was cited in several subsequent publications. This unconfirmed observation was sometimes even considered as a variant feature of fission yeast. Similar observations have been reported in localization dependence studies performed in other organisms as well (Cheeseman et al., 2004; Przewloka et al., 2007). These questions should be readdressed with the help of more sensitive assays in uniform experimental conditions in a variety of model systems. The outcome of these experiments will help us to precisely compare and contrast the KT structure and its function across species. The contrasting results of an identical question can occur due to the differences in experimental conditions or measurement techniques. For an example, localization dependence of Dsn1 on Mtw1 in S. cerevisiae is contradictory in two reports (De Wulf et al., 2003; Pinsky et al., 2003). More quantitative assays to determine the actual scenario are required in such cases to resolve these apparent discrepancies.
It is evident that although most of the proteins assemble at the CEN are functionally conserved across species, the CEN DNA is diverged even in closely related species. Comparative genomic analyses in different yeasts revealed that the CEN DNA is hyper-variable even in closely related species (Bensasson et al., 2008; Padmanabhan et al., 2008; Rhind et al., 2011). The phenomenon of hyper-variability of the DNA sequence at the CEN despite its conserved function in chromosome segregation was previously designated as the ‘centromere paradox’ (Henikoff et al., 2001).
In this review, we analysed the similarities and differences in the process of KT assembly in yeasts. While the organization of a KT is conserved, there appears to be subtle divergence in regulation of KT assembly in these organisms. Whether this process has evolved uniquely in different organisms to keep pace with the fast evolving CEN DNA is not clear. If the process of KT assembly occurs in a step-wise manner, a master regulator may control the recruitment of most of the KT proteins. In fact, an inner KT protein Ndc10 plays the central role in S. cerevisiae (Fig. 2a), while the middle KT proteins – Mis6 and Spc7 – play governing roles to a great extent in S. pombe (Fig. 2b). This process is remarkably diverged with a complex interdependence among many essential KT proteins from various layers in C. albicans (Fig. 2c). Unravelling this fascinating molecular mechanism of KT assembly in many organisms will improve our understanding of how the KT assembly pathways coevolved with the CEN DNA during speciation.
We thank B. Suma (Central instrumentation facility, Molecular Biology and Genetics Unit, Jawaharlal Nehru Centre for Advanced Scientific Research) for confocal microscopy and image processing. We are thankful to the members of Sanyal laboratory for insightful comments. We express our regret to our colleagues whose work could not be cited due to space limitations.