Keywords

16.1 Introduction

16.1.1 The SAC-KMN axis

The accurate and timely segregation of chromosomes during mitosis requires the formation of a bipolar mitotic spindle with stably attached chromosomes. Once all of the chromosomes are aligned properly, the connection between the sister chromatids is severed by the action of separase, a cysteine protease. Separase also contributes to centriole disengagement at the end of mitosis. Temporal and spatial coordination of these two activities with the rest of the cell cycle is required for the successful completion of mitosis. The accurate segregation of chromosomes when cells divide is ensured by the spindle assembly checkpoint (SAC) , a highly intricate regulatory mechanism that monitors and corrects defects in chromosome attachment to the metaphase plate. The SAC is controlled by a panel of protein kinases that includes Bub1, BubR1, Mps1 and Aurora B and the non-kinase proteins Mad1, Mad2, Bub3, BuGZ and Cdc20.

16.1.2 Multidomain Protein Kinases Regulate the SAC

Budding uninhibited by benzimidazoles 1 (Bub1), Budding uninhibited by benzimidazoles related 1 (BubR1), dual-specificity kinase Monopolar spindle 1 (Mps1) and Aurora B are multidomain serine/threonine protein kinases with essential roles in the SAC signalling pathway in higher organisms (Krenn and Musacchio 2015; Elowe 2011; Musacchio 2011; Boyarchuk et al. 2007; Abrieu et al. 2001). These SAC proteins have evolutionarily conserved catalytic domains in organisms ranging from budding and fission yeasts to worms to humans (Bavetsias and Linardopoulos 2015; Bolanos-Garcia and Blundell 2011). For instance, Bub1 is required for the proper assembly of the inner centromere (Boyarchuk et al. 2007). Phosphorylation of the kinetochore organiser protein KNL1 by Mps1 is required for the recruitment of Bub1, BubR1 and Bub3 to the kinetochore while Mps1 from fission yeast (known as Mph1 in this specie) phosphorylates Mad3 to inhibit Cdc20 (known as Slp1 in fission yeast) and this post-translational modification appears important to maintain SAC arrest (Zich et al. 2016). It has been proposed that Mps1 compete with microtubules to bind at kinetochores and that such competitive binding contributes to regulate SAC signalling (Hiruma et al. 2015; Ji et al. 2015). Whether additional interactions mediated by Mps1, microtubules and/or the Ndc80 complex are involved in this process is an aspect that remains to be clarified (Aravamudhan et al. 2015; Krenn and Musacchio 2015; Nilsson 2015).

The proteins Mitotic arrest deficient 1 and 2 (Mad1 and Mad2, respectively); Budding uninhibited by benzimidazoles 3 (Bub3); Cell division cycle protein 20 (Cdc20); Bub3-interacting GLEBS-motif-containing ZNF207 (BuGZ); and Rod, Zwilch and ZW10, which define the RZZ-complex (Lara-Gonzalez et al. 2012), are also central components of the SAC. BubR1 (known as Mad3 in yeasts) interacts with Bub3, Mad2 and Cdc20 to form the Mitotic Checkpoint Complex (MCC; see Fig. 16.1), which inhibits the Anaphase Promoting Complex/Cyclosome (APC/C) to prevent metaphase-anaphase transition (Zhang et al. 2016; Musacchio 2011; Chao et al. 2012). Some of the key interactions underpinning SAC signalling are presented in Table 16.1. It has been shown that in both mammalian cells and in the fission yeast Schizosaccharomyces pombe, kinetochores shortened after microtubule severing. However, whereas in fission yeast all kinetochores could relax to a similar length, in human cells the more stretched kinetochores remained more stretched suggesting that the differences are due to the increased structural complexity of the mammalian kinetochore (Cojoc et al. 2016).

Fig. 16.1
figure 1

A, crystal structure of a ternary heterotrimer of the Mitotic Checkpoint Complex (MCC) from Schizosaccharomyces pombe (pdb id 4AEZ). B, crystal structure of Cdc20 (pdb id 4GGA), another WD 40 fold protein that forms part of the MCC. Figures generated with PyMOL (DeLano 2002)

Full size image
Table 16.1 Summary of key SAC protein interactions that regulate cell cycle progression
Full size table

The protein BuGZ was identified recently as a critical component of the SAC that contributes to the targeting of the Bub1-Bub3 and BubR1-Bub3 heterodimer complexes to the kinetochore . BuGZ is predicted to contain a N-terminal zinc finger domain; a GLEBS motif that is conserved from yeast to human and essential to bind Bub3, and a region of low structural complexity (Jiang et al. 2014; Toledo et al. 2014). BuGZ binds to microtubules and tubulin to regulate the SAC in a process that involves BuGZ phase transition (coacervation) (Jiang et al. 2015). It will be important to clarify if coacervation is a physicochemical feature shared by other proteins that regulate spindle assembly. Ultimately, this knowledge can be used to define new cancer therapies that rely on mitosis inhibition (Herman et al. 2015).

Zeste-white 10 (ZW10) and rough deal (ROD) were initially identified in D. melanogaster. Both proteins are highly conserved among multicellular eukaryotes (Karess 2005; Scaërou et al. 1999, 2001; Williams et al. 1992; Karess and Glover 1989). Null mutations in the zw10 and rod genes and depletion of these proteins in C. elegans and vertebrate cells result in chromosome segregation defects and extensive aneuploidy in mitotic and meiotic cells. Co-immunoprecipitation of ZW10 and ROD in fly and human cell extracts and their immunostaining in mitotic cells suggest an interdependent recruitment of the proteins to the kinetochore (Williams et al. 1992). Zwilch mutations cause a similar mitotic phenotype as ROD and ZW10, thus confirming a role of this protein in the SAC. Zwilch, ROD and ZW10 from human extracts isolated by affinity chromatography methods using ZW10 as bait show the three proteins associate to form a stable complex that seems to contain two copies of each protein (Civril et al., 2010; Kops et al. 2005a; Williams et al. 2003). The recent crystallisation and preliminary X-ray crystallographic analysis of a human ROD-ZW10-Zwilch complex (Altenfeld et al. 2015) further supports the notion of a complex of the three constitutive subunits with a 2:2:2 stoichiometry.

The control of chromosome segregation in higher organisms requires communication of SAC sub-complexes with the KMN (KNL1/Mis12/Ndc80) network, a multiprotein macromolecular assembly that constitutes the structural core of the kinetochore and is essential for the establishment of proper kinetochore-microtubule attachments (Liu et al. 2016; Aravamudhan et al. 2015; Ghongane et al. 2014; Santaguida and Musacchio 2009). Recently considerable progress has been made in understanding the composition of the kinetochore, the recruitment hierarchy of its components, and the principles of its regulation (revised in Agarwal and Varma 2014; Przewloka and Glover 2009). However, structural details of the molecular interactions of the KMN network have proved elusive even though they are clearly indispensable for the mechanism of kinetochore assembly and SAC signalling. The understanding of the function of the individual subunits and the different sub-complexes of the KMN network requires a description of the functional and structural features of its central components, and this is presented below.

16.2 The Kinetochore Null Mutant 1 (KNL1)

The kinetochore protein KNL1 (also known as CASC5, AF15q14 and Blinkin in humans; Spc105 in budding yeast, Spc7 in fission yeast, and KNL1 in humans and C. elegans) acts as a multi-substrate docking platform of the KMN network . KNL1 was initially identified in Saccharomyces cerevisiae as component of the spindle pole body (hence the acronym Spc105) (Nekrasov et al. 2003), which in C. elegans is commonly referred to as KNL1 (kinetochore-null phenotype 1) (Cheeseman et al. 2004) and Spc105R (Spc105-related) in Drosophila (revised in Przewloka and Glover 2009). Spc105R from Drosophila shows considerable sequence divergence compared to other species. This feature of the KMN contributes to define a distinctive structural organisation to the KMN network of the fruit fly. Depletion of KNL1 of higher organisms by RNAi causes severe chromosomal segregation defects that resemble phenotypes observed after depletion of the SAC kinases Bub1 and BubR1, including premature exit from mitosis and early onset of anaphase (Przewloka and Glover 2009). KNL1 is the largest subunit of the KMN network and is required for accurate chromosome segregation during mitosis (Desai et al. 2003). KNL1 integrates SAC kinase and phosphatase activities and contributes to the formation of kinetochore-microtubule attachments (Przewloka and Glover 2009; Santaguida and Musacchio 2009). Evidence of the precise role of functional regions in KNL1, including the motifs SILK, RVSF, MELT, KI, and a domain that adopts the RWD fold is providing new insights how KNL1 coordinates SAC activity. This is an important aspect of the SAC that has been revised recently (Caldas and DeLuca 2014; Ghongane et al. 2014).

16.3 The Ndc80 Complex

Microtubules contribute to key biological processes controlled by cell motility, including the maintenance of cell orientation and the regulation of focal adhesion turnover (Alushin et al. 2010). In human cells the Ndc80 complex is linked to centromeric chromatin to mediate end-on attachment of spindle microtubules in a process that requires Ndc80 binding to the kinetochore proteins CENP-T and CENP-C (Suzuki et al. 2015; Tanaka 2013). In humans, the Ndc80 complex is composed of the proteins Hec1 (a subunit that is also commonly referred to as Ndc80), Nuf2, Spc24 and Spc25 (DeLuca and Musacchio 2012; Tooley and Stukenberg 2011; Varma and Salmon 2012). The Ndc80 complex adopts a dumbbell-shape architecture with the subcomplexes Nuf2-Ndc80 and Spc24-Spc25 located in opposite ends of the molecule (see Fig. 16.2) (Ciferri et al. 2005; Wei et al. 2005). In one hand, Nuf2 binding to Ndc80 is required for the localisation of the Ndc80 complex to microtubules. On the other hand, Spc24-Spc25 heterodimer complex formation appears to be a pre-requisite for the binding of Spc24-Spc25 to KNL1 and the Mis12 complex (Cheeseman et al. 2006; Kiyomitsu et al. 2007; Wei et al. 2007; Ciferri et al. 2008; Wan et al. 2009; Joglekar and DeLuca 2009). The hairpin region of Ndc80 seems to be an important structural requirement for the effective kinetochore recruitment of Mps1 (Mph1), at least in fission yeast (Chmielewska et al. 2016).

Fig. 16.2
figure 2

Crystal structure of a chimeric (bonsai) Ndc80 complex (pdb id 2VE7) revealed a dumbbell-shape topology. The sub-complexes Nuf2-Ndc80 and Spc24-Spc25 are located in opposite ends of the molecule

Full size image

Protein depletion assays coupled to quantification of kinetochore protein copy numbers in human cells have provided clues of the extent of the interactions between centromeric chromatin and the microtubule-binding Ndc80 complex. Such studies revealed about 244 Ndc80 complexes per human kinetochore (i.e., approximately 14 per kinetochore microtubule) and 151 Ndc80 complexes are associated to the KMN network . These studies also showed that each CENP-T molecule recruited approximately two copies of the Ndc80 complex (one as part of a KMN network) with nearly 40% of CENP-C recruited exclusively to the KMN network (Suzuki et al. 2015). Undoubtedly, the quantification of kinetochore protein copy numbers in other species should provide important new insights into the evolution and the subtle differences in the mode of regulation of the KMN network.

16.4 Mis12 Complex

In humans, the Mis12 complex (also known as the MIND complex) consists of the proteins Mis12, Dsn1, Nsl1 and Nnf1. The complex is a central component of the KMN network . Nsl1 has been identified as a link between the human Mis12 and Ndc80 complexes (Petrovic et al. 2010) and this protein seems to play a similar important role in yeast kinetochores (Kudalkar et al. 2015). Indeed, recent studies in yeast cells have shown that the interaction between the Mis12 complex (which in yeast is made up of the proteins Mtw1, Nsl1, Nnf1, Dsn1) and the Ndc80 complex (constituted by the proteins Ndc80, Nuf2, Spc24, Spc25) (Biggins 2013) is mediated by an extensive number of contacts (Kudalkar et al. 2015). Mis12 alone does not bind to microtubules but it does bind to the Ndc80 complex, an interaction that enhances Ndc80 binding to microtubules (Kudalkar et al. 2015). In Drosophila, the Mis12 complex localises to the mitotic centromere in a process that implicates the binding of the N-terminal regions of Mis12 and Nnf1 to CENP-C (Richter et al. 2016). Reconstitution of the yeast Mis12-Ndc80 assembly coupled to cross-linking analysis revealed an intricate set of interactions involving five of the eight proteins within the Mis12 and Ndc80 complexes and a direct interaction between the proteins Nsl1 and Spc24/Spc25. The latter interaction defines a unique interface thus suggesting that in different organisms the regulation of Ndc80 functions may be achieved by a distinctive mode of interactions (Kudalkar et al. 2015).

16.5 Centromere-Associated Protein E (CENP-E)

CENP-E is required to maintain a stable genome through the stabilisation of microtubule capture in the kinetochore . CENP-E functions as a highly processive plus end-directed motor that couples chromosome position with microtubule depolymerisation thus linking kinetochores to dynamic spindle microtubules. CENP-E participates in the recruitment of BubR1, Mad1 and Mad2 to attached and newly unattached kinetochores and its binding to the SKAP protein is required for accurate chromosome segregation in mitotic cells (Huang et al. 2012). SKAP seems to form part of the kinetochore corona fibres of mammalian centromeres as judged by immunoelectron microscope imaging. The interaction between CENP-E and SKAP, which involves the C-terminal tail of the former protein, is thought to be essential for kinetochore-microtubule attachment in vivo. Depletion of SKAP or CENP-E by RNA interference drastically reduces inter-kinetochore tension, thus leading to chromosome segregation defects and a prolonged delay to fulfill metaphase alignment (Huang et al. 2012).

In human cells CENP-E kinetochore localisation depends on its binding to Nuf2, an interaction that is mediated by C-terminal regions of both CENP-E and Nuf2 as determined with the yeast two-hybrid system and pulldown assays (Liu et al. 2007). Moreover, depletion of human Nuf2 by small interfering RNA abolished CENP-E kinetochore localisation and resulted in chromosome segregation defects, thus confirming the requirement of Nuf2 for CENP-E localisation to the kinetochore and the essential role of the interaction for the correct segregation of chromosomes in mitosis (Liu et al. 2007).

16.6 The Centromeric Nucleosome-Associated Network (CCAN)

Centromeres are differentiated chromatin domains, present once per chromosome, that direct segregation of the genome in mitosis and meiosis by specifying assembly of the kinetochore . The latter provides an essential link that brings together chromosomes and spindle microtubules (Pesenti et al. 2016; Rago et al. 2015). The specific spatial configuration of the centromere is likely to contribute to the tight regulation of mitosis and the dynamics of kinetochore-microtubule attachments (George and Walworth 2016). The centromeric nucleosome-associated network (CCAN) is a constitutive complex that is assembled onto centromeric CENP-A chromatin and widely considered as the prime candidate for specifying centromere identity (Foltz et al. 2006). The CCAN is composed by the proteins CENP-C, CENP-H/ CENP-I/ CENP-K, CENP-L/ CENP-M/ CENP-N, CENP-O/CENP-P/CENP-Q/CENP-R/CENP-U, CENP-T/CENP-W, and CENP-S/CENP-X (Foltz et al. 2006; revised by Perpelescu and Fukagawa 2011). The CCAN recruits the outer kinetochore components of the KMN network KNL1, the Mis12 complex, and the Ndc80 complex thus bringing together kinetochore proteins and spindle microtubules. Disruption of the interaction between CENP-A and CCAN causes errors of chromosome alignment and segregation that prevent cell survival (Foltz et al. 2006).

CENP-A is a centromere-specific isoform of histone H3 (Perspelescu and Fukugawa 2011; Stoler et al. 1995; Palmer et al. 1991) that contributes to kinetochore formation and centromere-kinetochore assembly thus guiding the movement of chromosomes and cell cycle progression throughout mitosis (Fachinetti et al. 2013; Mendiburo et al. 2011; Barnhart et al. 2011; Wan et al. 2009). The crystal structures of two human centromeric nucleosomes, one containing CENP-A (pdb id 3AN2) and one containing CENP-C in complex with the cognate α-satellite DNA derivative (pdb id 4X23) revealed that the latter molecule wraps around a histone octamer (see Fig. 16.3). Such octameric complex is defined by the assembly of two copies of histones H2A, H2B, H4 and CENP-A (Tachinawa et al. 2011).

Fig. 16.3
figure 3

(a) Crystal structure of a centromeric nucleosome in complex with CENP-A; (b) crystal structure of a centromeric nucleosome in complex with CENP-C. In both cartoons the view is in the axis of the DNA supercoil

Full size image

In addition to CENP-A, CENP-C and CENP-T contribute to kinetochore assembly in vertebrates as shown by studies in which the DNA-binding regions of CENP-C and CENP-T were replaced with alternate chromosome-targeting domains, thus resulting in the localisation of functional CENP-C and CENP-T to ectopic loci and a CENP-A-independent assembly of the kinetochore (Gascoigne et al. 2011). Furthermore, phosphorylation of CENP-T appears as an important requirement for proper mitotic assembly of both endogenous and ectopic kinetochores (Gascoigne et al. 2011).

CENP-H is an inner kinetochore protein that is highly conserved amongst eukaryotes (Orthaus et al. 2006). CENP-H directly interacts with CENP-K through multiple contacts to form a stable heterodimeric complex. CENP-H and CENP-K are predicted to contain extensive coiled-coil regions that seem to play an important role in the stabilisation of the CENP-H-CENP-K heterocomplex (Qiu et al. 2009). Depletion of CENP-H in human cells led to severe mitotic phenotypes including misaligned chromosomes and multipolar spindles but not mitotic arrest (Orthaus et al. 2006). CENP-H depletion results in reduced levels of CENP-E but only slightly affects the levels of CENP-C bound to the kinetochore while suppression of CENP-H expression has not effect on BubR1 kinetochore localisation and a SAC response (Orthaus et al. 2006).

The CENP-T/W complex assembles in late S and G2 phases of the cell cycle and is required for mitosis. The CENP-T/W complex is integrated with centromeric chromatin in association with Histone H3 nucleosomes (Prendergast et al. 2011) but it does not persist across cell generations. Instead, association of H3 with the CENP-T/W complex seems to be specific for the regulation of kinetochore activity (Prendergast et al. 2011). CENP-T centromere localisation is restricted to the S-phase of the cell cycle. CENP-T directly associates with CENP-A and CENP-B as shown by Förster resonance energy transfer (FRET) studies. Taken together these studies indicate that CENP-T is required for the recruitment of other proteins to the kinetochore (Hellwig et al., 2008). Furthermore, centromeric-bound CENP-T-W and CENP-S-X subcomplexes associate to form a stable CENP-T-W-S-X heterotetramer that binds to DNA to form supercoil structures (Takeuchi et al. 2014; Nishino et al. 2012). High-resolution structural analyses of the individual subcomplexes and the tetramer have revealed important structural similarities with the nucleosome and certain histone fold-containing complexes (Nishino et al. 2012).

In human cells the inner kinetochore components CENP-C and CENP-T function in parallel pathways to recruit the KMN network to the kinetochore (Nishino et al. 2013; Schleiffer et al. 2012) as shown by independent ectopic targeting of these proteins to a chromosomal locus (Rago et al. 2015). For instance, the physical interaction of CENP-C with KNL1 and the Mis12 complex is required for the recruitment of the Ndc80 complex to the kinetochore whereas CENP-T kinetochore recruitment is only dependent of CENP-T binding to the Ndc80 complex. Furthermore, the CENP-T-Ndc80 complex assembly in turn promotes KNL1/Mis12 complex recruitment in a process that implicates a separate region on CENP-T (Rago et al. 2015). The formation of the CENP-C and CENP-T sub-complexes seems to obey different regulatory controls: the recruitment of the KMN network to CENP-C is stimulated by Aurora B kinase while that of CENP-T is regulated by cyclin-dependent kinase (Cdk) (Rago et al. 2015).

A number of additional microtubule plus-end binding proteins that have been associated with the kinetochore include the CLIP-associating protein 1 (CLASP-1) and 2 (CLASP-2), Astrin, Kinastrin, KIF2B, Kif18A and SKAP. The general structural and functional features of these proteins and their roles in SAC signalling are described below.

16.7 CLASP-1 and CLASP-2

The microtubule plus-end binding proteins CLASP-1 and CLASP-2 play important roles in the regulation of the density, length distribution and stability of interphase microtubules thus integrating spindle and kinetochore functions (Pereira et al. 2006; Maiato et al. 2003). In yeast, Drosophila, and Xenopus, one CLASP orthologue is present, whereas in human two proteins have been identified: CLASP-1 and CLASP-2. In all these organisms CLASP proteins are required for mitotic spindle assembly through the regulation of microtubule dynamics at the kinetochore. In mitotic cells both proteins associate with the ends of growing microtubules and with kinetochores in a process that requires the binding of these proteins to EB1 (Mimori-Kiyosue et al. 2005). The interaction of CLASP-1 and CLASP-2 with EB1 implicates the middle region of both CLASPs (Mimori-Kiyosue et al. 2005). At least in HeLa cells CLASP-1 and CLASP-2 show similar and at least partially redundant roles in organising the mitotic apparatus (Pereira et al. 2006). Their simultaneous depletion results in extensive mitotic spindle defects and an abnormal exit from mitosis. Targeting CLASP-1 with specific anti-CLASP-1 antibodies impairs microtubule dynamics in the kinetochore and the mitotic spindle, leading to the formation of abnormal monopolar asters in which the chromosomes are found buried in the interior. Similarly, the expression of a truncated form of CLASP-1 lacking the kinetochore binding domain results in the formation of depolymerisation-resistant microtubule bundles with a radial array (Maiato et al. 2003). Inhibition of glycogen synthase kinase -3 (GSK3) activity by the tyrosine kinase receptor ErbB2 regulates microtubule capture and stabilisation. Inhibition of Glycogen synthase kinase 3 beta (GSK3b) causes relocalisation of CLASP-2 to the plasma membrane and ruffles (Zaoui et al. 2010). All these observations strongly support an important role for CLASP-1 and CLASP-2 in the organisation of the mitotic spindle and the control of microtubules attachments.

It has been suggested that microtubules in vertebrate somatic cells are not only formed by the centrosome but that a significant number of them originate from the Golgi apparatus in a centrosome-independent manner. The process requires CLASPs recruitment to the trans-Golgi network by the protein GCC185 (Efimov et al. 2007; Zhonghua et al. 2007). However, mechanistic details of the regulation of spindle assembly in mitosis by membrane systems remain largely obscure.

16.8 Astrin and Kinastrin

Astrin is a mitotic spindle-associated protein found in most human cell lines and tissues that is required for proper chromosome alignment at the metaphase plate; is essential for progression through mitosis and contributes to the regulation of separase activity (Dunsch et al. 2011; Thein et al. 2007; Gruber et al. 2002). Depletion of this protein by RNA interference delays chromosome alignment, leads to the loss of spindle architecture and sister chromatid cohesion before the onset of anaphase, and ultimately results in apoptosis (Gruber et al. 2002). Amino acid sequence analysis and fold recognition bioinformatics tools suggest that Astrin has an N-terminal globular domain and an extended coiled-coil domain. Electron microscopy studies of recombinant Astrin showed that this protein self-associates to form parallel dimers with head-stalk structures reminiscent of motor proteins. However, the low amino acid sequence identity and structural similarity to known motor proteins requires further investigations to establish to what extent there is a functional correspondence between Astrin and kinesins.

Kinastrin is the major interacting partner of Astrin in mitotic cells and the interaction is required for Astrin targeting to microtubule plus ends. Overexpression or depletion of Kinastrin mislocalise Astrin and causes mitotic defects that resemble those observed in Astrin-depleted cells. Astrin and Kinastrin can form a complex with SKAP, which also co-localises to microtubule plus ends to facilitate chromosome alignment (Dunsch et al. 2011). These observations support the notion that the microtubule plus end targeting activity of Astrin is required to sustain spindle architecture and to ensure chromosome alignment and that perturbation of these interactions delay mitosis and cause the premature activation of separase (Dunsch et al. 2011). Interestingly, Astrin acts as a negative regulator of mTORC1, which seems to be essential to elicit a cellular stress response. Under stress conditions, Astrin blocks mTORC1 self-association and recruits Raptor, a protein component of mTORC1, to stress granules, thus preventing apoptosis caused by the induction of mTORC1 hyperactivation (Thedieck et al. 2013). This is an exciting finding that suggests a potential link between cellular stress response and the control of chromosome segregation. Further studies should aim to clarify this aspect of SAC signaling.

16.9 KIF2B and Kif18A

The human genome has three genes (Kif2a, Kif2b, and MCAK [also known as Kif2c]) that encode for kinesin-13 proteins. Kif2a, Kif2b, and MCAK fulfill distinct functions during mitosis in human cells (Hood et al. 2012; Manning et al. 2007). Human kinesin Kif18A is a kinesin-8 protein and microtubule-depolymerising protein that contributes to stabilise the CENP-E-Bub1 complex at the kinetochores during early mitosis (Mayr et al. 2007). In vitro, Kif18A shows a slow plus-end-directed microtubule depolymerising activity whereas in mitotic cells in vivo Kif18A localises close to the plus ends of kinetochore microtubules. Depletion of Kif18A induces aberrant mitotic spindles and loss of tension across sister kinetochores and activates the SAC (Mayr et al. 2007). During vertebrate cell division, chromosomes oscillate with periods of smooth motion and rapid reversals in direction. These fluctuations must be spatially constrained to ensure the proper alignment and high fidelity segregation of chromosomes. In humans, Kif18A plays an essential role in the control of chromosome oscillations by reducing the amplitude of pre-anaphase oscillations and slowing down poleward movements during anaphase. This manner, Kif18A contributes to the control of kinetochore microtubule dynamics underlying chromosome positioning in mitosis (Gardner et al. 2008; Stumpff et al. 2008). Moreover, Kif18A physically interact with CENP-E and BubR1 during mitosis as revealed by co-immunoprecipitation studies. Kif18A depletion results in mitotic arrest and chromosome missalignment and stimulates CENP-E degradation indicating that chromosome congression defects due to Kif18A depletion are at least in part mediated through destabilisation of CENP-E (Huang et al. 2009).

16.10 SKAP

SKAP is an essential component of the mitotic spindle that associates with kinetochores and is required for chromosome alignment, normal timing of sister chromatid segregation and maintenance of spindle pole architecture (Fang et al. 2009). SKAP also plays a role in the control of kinetochore oscillations and the regulation of microtubule plus-ends dynamics during mitosis (Wang et al. 2012a). Although suppression of SKAP expression does not stimulate the SAC, it substantially increases the duration of metaphase, delays the activation of separase and decreases the fidelity of chromosome segregation (Fang et al. 2009).

SKAP binds to microtubules in vitro, an interaction that is synergised by CENP-E. Thus, CENP-E and SKAP work together to control dynamic kinetochore -microtubule interactions (Huang et al. 2012). SKAP binds to the C-terminal tail of CENP-E in vitro and is essential for an accurate kinetochore-microtubule attachment in vivo. Depletion of SKAP or CENP-E by RNA interference drastically impairs inter-kinetochore tension and causes chromosome missegregation (Wang et al. 2012 b; Huang et al. 2012). SKAP also interacts with Mis13, which seems important for the accurate interaction between kinetochore and dynamic spindle microtubules. SKAP directly binds Mis13 and the interaction specifies the kinetochore localisation of the former protein, an observation that has been confirmed by small interfering RNA studies to suppress Mis13 expression (Wang et al. 2012b). A complex formed between SKAP and Astrin-Kinastrin localises to microtubule plus ends to facilitate proper chromosome alignment (Dunsch et al. 2011). Further studies should aim to clarify the role of these interactions in the control of SAC signalling.

16.11 Disorder-to-Order Transitions in the KMN

It is worth noting that many of the kinetochore proteins described above are predicted to contain large regions of low structural complexity (see Fig. 16.4). A pattern of disorder-to-order transitions in SAC signalling has emerged from the structures of diverse complexes involving the kinetochore organiser protein KNL1, including the N-terminal TPR-containing domains of Bub1 and BubR1 in complex with the KNL1 N-terminal KI motifs; Bub3 bound to KNL1 MELT motifs and the KNL1 RWD domain in complex with a synthetic peptide that mimics the protein Nls1 (Bolanos-Garcia et al. 2011; Krenn et al. 2012; Primorac et al. 2013; Petrovic et al. 2014) (see Fig. 16.5). One distinctive feature that emerges from the analysis of the above mentioned complexes is the predominance of cooperative hydrophobic interactions that stabilise the complexes. With the exception of the C-terminal region which contains a globular RWD domain, multiple regions of low structure complexity that span most of the polypeptide chain occur in KNL1. This is not surprising because multiple regions of low structure complexity occur often in hub proteins that define interactome networks (Babu et al. 2012; Kim et al. 2006; Dosztányi et al. 2006; Dunker et al. 2005, 2008; Haynes et al. 2006). Indeed, in multiple biological systems cooperative interactions involve the recognition of a flexible protein by a globular one, leading to concerted folding and binding (Blundell et al. 2002). This is particularly evident in hub proteins that define interactome networks because such proteins contain intrinsic local disordered regions (Dunker et al., 1998; Gsponer and Babu, 2009) that often associate with interacting partners through concerted binding and folding (Uversky 2015; Dosztányi et al. 2006; Dunker et al. 2005). The general model for concerted folding upon binding appears to be initial binding of a large side chain into a deep pocket, usually followed by interaction at a second and sometimes third pocket, forming a cluster of small pockets (Fuller et al. 2009). Less conserved interactions involving regions N- or C-terminal to the conserved motif then fold cooperatively onto the surface of the globular partner. There are examples of this type of interactions in SAC signalling, including the binding of the KI motif of KNL1 to BubR1 (Bolanos-Garcia et al. 2011) and possibly that of an equivalent KI motif in KNL1 specific to Bub1 (Krenn et al. 2012) and the binding of Bub3 to the MELT motifs of KNL1, a sequential, multisite interaction that is subjected to phospho-regulation (Vleugel et al. 2015). Indeed, the interplay of phosphorylation and dephosphorylation cascades rises as an important mechanism to regulate the SAC (Manic et al. 2017; Funabiki and Wynne 2013; London et al. 2012; Shepperd et al. 2012; Rosenberg et al. 2011; Liu et al. 2010).

Fig. 16.4
figure 4

Different KNL1 complexes revealed a similar mode of binding underlying disorder-to-order transitions . A, crystal structure of human TPR BubR1 in complex with the KNL1 KI motif (pdb id 3SI5); crystal structure of a Bub3-Bub1 GLEBS motif-KNL1 MELT motif complex (pdb id 4BL0). C, KNL1 RWD domain in complex with a Nsl1 fragment (pdb id 4NF9).

Full size image
Fig. 16.5
figure 5

The presence of large segments of disordered regions in multiple kinetochore binding proteins is a common structural feature. The plot shows the disorder predictions based on PONDR-FIT (Xue et al. 2010) analyses.

Full size image

Furthermore, the reciprocal communication of disorder-to-order transitions on two or more distant functional surfaces of high intrinsic disorder can maximise allosteric coupling between proteins. This mode of molecular recognition and signal amplification may obey the same mechano-chemical principles underlying the interaction of simpler systems such as binding of a biotin repressor to biotin protein ligase (Egington et al. 2015). Also, macromolecular crowding effects (Mourão et al. 2014; Cino et al. 2012; Babu et al. 2012; Wang et al. 2012a) can be anticipated to play a major role in the regulation of the SAC given the prominent role of proteins with multiple regions of low structural complexity in the process, including kinesin motors (Leduc et al. 2012).

In summary, regulation of the rate in which spindle microtubules attach/detach to/from kinetochores plays a central role in the control of chromosome segregation. Multiple mechanisms of assembly and holistic models that take into account the role of protein receptors, signalling networks and regulatory feedback mechanisms have been proposed in an attempt to describe more precisely the role of kinetochore -microtubules interactions for the control mitotic progression in higher organisms (Kim and Yu 2015; Godek et al. 2015). As discussed below, disruption of this balance quickly results in aneuploidy , genome instability , cancer and diverse birth and development defects.

16.12 The SAC-KMN Axis in Disease

Mistakes in the process of cell division can lead to the rearrangement, the loss or gain of chromosomes (aneuploidy). Solid tumors are frequently aneuploid, and many display high rates of chromosome missegregation and chromosomal instability (CIN). The most common cause of CIN is the persistence of aberrant kinetochore -microtubule attachments, which manifest as lagging chromosomes in anaphase. Errors in kinetochore-microtubule attachments during prometaphase can be due to stochastic interactions between kinetochores and microtubules.

Mps1 has been identified in the signature of the top 25 genes overexpressed in CIN and aneuploid tumours (Kops et al. 2005b; Carter et al. 2006) and found to be upregulated in a number of tumours of different origins including bladder, anaplastic thyroid, breast, lung, esophagus, and prostate. In the absence of a functional mitotic checkpoint, as occurs when Mps1 function is lost, cells become rapidly aneuploid and subsequently die (Kops et al. 2005b; Janssen et al. 2009). This feature, together with the observations that inhibiting Mps1 with chemical inhibitors kills cultured tumour cells (Kwiatkowski et al. 2010) and that even its partial inhibition creates tumour cells more sensitive to clinical doses of taxol (Janssen et al. 2009), show that targeting Mps1 with drugs may be beneficial to arrest proliferation of tumour cells. Significant chromosomal abnormalities are the cause of severe diseases such as breast cancer , the most common cancer in the UK. This year alone, 50,000 people in the UK will find out they have breast cancer and 12,000 people will die from it (Cancer Research UK organisation). The fact that Mps1 inhibition in tumour xenograft models significantly reduces tumour growth rates while leaving normal cell growth unaffected (Daniel et al. 2011) makes Mps1 an attractive target for cancer therapy (Kapanidou and Bolanos-Garcia 2014). In addition to Mps1, Aurora B kinase is the cellular target of diverse Medicinal Chemistry campaigns to develop inhibitors that function as adenosine triphosphate (ATP) competitors of these protein kinases. Also important is the search of ubiquitin ligase inhibitors that target the E3 ubiquitin ligase activity of the APC/C complex and APC/C regulators (Zhang et al. 2014, 2016; Zhou et al. 2013, 2016; Fujimitsu et al. 2016).It can be anticipated a steady increase of activity in this area in the coming years.

Defects in centrosome and spindle-associated functionsare the most frequent cause of primary microcephaly syndromes in humans. For example, mutations in CENP-E have defined a novel kinetochore -centromeric mechanism for microcephalic primordial dwarfism (Mirzaa et al. 2014) while centromere protein F (CENP-F) has been implicated in Hutchinson-Gilford progeria syndrome, a rare disorder that leads to premature ageing and death due to myocardial infarction or stroke. The disease is caused by expression of the protein Progerin, which is a truncated version of the protein prelamin A (Eisch et al. 2016). Progerin displaces CENP-F from metaphase chromosome kinetochores, thus increasing chromatin lagging and causing genome instability (Eisch et al. 2016).

The range of malignancies described above indicate that the development of new drugs to interfere with abnormal cell proliferation is urgently required. The development of new drugs to interfere with defective SAC signalling and its communication with the KMN network in human tumours appears as an attractive alternative to prevent the proliferation of cells carrying abnormalities in chromosome structure and number. Multiple protein-protein interactions in regulatory hubs that control chromosome segregation and mitosis progression may constitute an important pool of novel drug targets . Structural insight into the molecular architecture of key interactions that regulate the SAC should pave the way for drug target identification and validation. In the absence of high resolution structural data, definition of the relationship between hub proteins and drug targets based on the combinatorial analysis of intrinsic structural disorder and gene onthology seems particularly attractive (Fu et al. 2015; The Gene Ontology Consortium 2010).

16.13 Emerging Methods in Structural Biology

The dynamic and coordinated assembly and disassembly of protein complexes in time and space follows sequential obligate stages that result in an enhanced selectivity with a low margin for errors in the process. At the same time, the dynamics of protein complex assembly and disassembly represents a great challenge for their structural and functional characterisation and often requires a combinatorial multidisciplinary approach involving a range of biochemical, biophysical, molecular and cellular approaches. Recent advances in Förster resonance energy transfer by fluorescence lifetime imaging microscopy; laser ablation; small angle x-ray scattering in structural biology (SAXS); nuclear magnetic resonance (NMR); serial femtosecond crystallography; and cryo-transmission electron microscopy (TEM) represent new exciting opportunities to understand the complex dynamics and mode of regulation of the SAC-KMN-microtubule signaling axis through the combinatorial use of the techniques.

Advances in SAXS methods allow the study of macromolecular complexes in solution that provide information about the shapes, conformations; oligomeric states of globular, non-globular and disordered macromolecules (Chaudhuri 2015) whereas multinuclear relaxation dispersion NMR methods permit to follow molecular recognition events of intrinsically disordered proteins in solution (Schneider et al. 2015; Parigi et al. 2014). Time-resolved protein crystallography using ultrafast X-ray free-electron lasers (XFELs) make it possible to follow rapid structural changes resulting from photolysis in the crystalline state and to resolve reaction intermediates at impressive high resolution (Barends et al. 2015; Tenboer et al. 2014). More recent improvements in serial femtosecond crystallography allowed the collection of X-ray diffraction patterns using X-ray pulses of 50 femtosecond duration that contained approximately 2 × 1012 photons per pulse to achieve a high-resolution XFEL structure of 1.75 Å (Ginn et al. 2015). Laser ablation has been used recently to separate microtubules attached to a merotelic kinetochore to study the mechanical response of the kinetochore resulting from changes of its length (Cojoc et al. 2016). At the same time, the study shows that the use of merotelic kinetochores emerges as an attractive experimental model for studying the mechanical properties of the kinetochore in live cells (Cojoc et al. 2016).

Equally impressive is the pace of instrumental and computational improvements in TEM where modern electron microscopes can produce images at a resolution higher than 2.0 Å (Glaeser 2016; Nogales 2016). A big gain that TEM offers is the possibility of studying samples of a relatively heterogeneous nature, thus allowing the analysis of multiple structural states that recapitulate the dynamics of complex protein-protein interactions including their mode of regulation and assembly/disassembly under diverse conditions (Weis et al. 2015; Louder et al. 2016). Parallel advances in cryo-electron tomography now allow the visualisation of macromolecular assemblies of irregular shapes; of organelles and even entire cells at the subnanometre resolution scale. An excellent review on this topic has been reported recently by Helen Saibil’s group (2016).

16.14 Closing Remarks

The study of cell division, the mechanism of transmission of the genetic material to descendants and the molecular basis of premature aging and cancer are areas of great interest in the Biomedical Sciences. Spindle assembly checkpoint (SAC) signalling is a truly fundamental cellular process of higher organisms that ensures the faithful segregation of chromosomes each time a cell divides. Undoubtedly, the inhibition of aberrant SAC signalling will benefit a wide range of disciplines, ranging from the cellular and molecular understanding of cell division in health and disease to the study of cell development , genome stability, ageing and comparative genomics.

The synergistic combination of biochemical, biophysical and structural biology methods for the characterisation of dynamic macromolecular complexes together with cellular and systems biology approaches should lead to a more comprehensive understanding of the cell and provide insights into how defects in molecular interactions can lead to the impairment of cellular regulation and function.

As the use of these experimental techniques alongside with molecular and computational methods begin to give insights into the dynamics of protein assembly /disassembly and their architecture, we will learn more mechanistic details of the remarkable complexity of the network of interactions between thousands of protein components that regulate metabolic and signalling pathways essential to all eukaryotes.

Because large multi-protein complexes play critical roles in cell regulation, interfering with the dynamics of their assembly and/or dissociation rises as an attractive strategy for the treatment of diseases. Extending the study of the structure and the dynamics of isolated SAC-KMN-microtubule sub-complexes to the molecular understanding of the mode of organisation of larger assemblies that ensure signal generation and amplification in a narrow spatial-temporal framework continues to represent a major challenge. Recent advances in TEM, electron-free lasers and a range of biophysical methods herald a new and exciting era for the molecular understanding of nuclear complexes that ensure genome stability to an unprecedented level of detail.