Spindle assembly and chromosome positioning in female meiosis

The majority of what is known about the spindle assembly bases on studies in mitotic cells. In mitosis, the spindle microtubules are nucleated predominantly by two centrosomes (see 1.1.2.1.). Presence of centrosomes in mitosis also pre-determines the polarity of the spindle. Capture of kinetochores by microtubule emanating from opposite poles is responsible for chromosome congression and bipolar orientation of sister kinetochores (see 1.1.3.3.).

The process of spindle assembly and chromosome positioning within the spindle in the absence of centrosomes is much less understood. Female meiosis is fundamentally different from mitosis. Primarily, the spindle lacks canonical centrosomes and thus relies on different mechanisms of spindle microtubule nucleation and organization (see 1.2.2.). Secondly, in the first round of meiotic division in females and males, sister kinetochores are attached to the same pole, while homologous chromosomes face opposite poles (see 1.3.2.1.).

In the absence of centrosomes, like in oocytes, chromosomes play a major role in spindle assembly (see 1.1.2.2.). After NEB, spindle microtubules nucleate in the vicinity of chromosomes and then, progressively, become organized into a bipolar structure. However, it is not well understood how chromosomes initiate spindle assembly in oocytes and how they establish contact with microtubules. Hence, it is also not known to what degree chromosomes contribute to establishment of bipolarity of the spindle in oocytes and how chromosomes become congressed and bi-oriented within the acentrosomal spindle.

1.3.1. Regulation of spindle assembly by chromosomes

In oocytes, chromosomes promote polymerization of spindle microtubule after NEB (see 1.2.1.). Chromosomes have been proposed to do so by modulating cytoplasmic gradients promoting microtubule assembly (see 1.1.2.2.). Similarly to in Xenopus egg extract, the presence of cytoplasmic conditions is sufficient to allow spindle formation in oocytes. Removal of chromosomes allowed spindle formation in mouse oocytes (Brunet et al., 1998). Overexpression of mutated form of Subito in Drosophila produced multiple spindles unassociated with chromosomes (Jang et al., 2007).

While in vitro studies provide compelling evidence for a role of a RanGTP gradient in chromosome-mediated spindle assembly, the role of RanGTP in vivo is much less understood (Dumont et al., 2007; Schuh and Ellenberg, 2007; Cesario and McKim, 2011). Decreasing a RanGTP level in mouse oocytes delays meiosis I spindle formation, but the formed spindles are functional, as they lead to correct chromosome segregation (Dumont et al., 2007). Spindle assembly in meiosis II is more affected and results in large spindle defects and formation of ectopic microtubule asters. Additionally, one of RanGTP main effectors, TPX2, is not present in mouse oocytes in prophase I and gradually accumulates later during meiosis (Brunet et al., 2008). Accordingly, depletion of TPX2 does not affect initial meiosis I spindle assembly, but leads to collapsing of the spindle later in prometaphase I. Therefore, RanGTP seems more important for later stages of spindle assembly in mouse female meiosis. Similarly, in Drosophila female meiosis I, decreasing a RanGTP level does not seem to affect the chromosome-mediated spindle microtubule nucleation and assembly (Cesario and McKim, 2011). However, depletion of Mei-38, a homolog of TPX2 in Drosophila, results in poorly organized spindles (Goshima et al., 2011; Wu et al., 2008).

In in vitro studies, CPC has been shown to stimulate chromosome-mediated spindle formation by locally inhibiting microtubule destabilizers (see 1.1.2.2.). A CPC component, Incenp, known localize the complex and thus target Aurora B function in mitosis, has been shown to be involved in acentrosomal spindle formation in Drosophila female meiosis I (see 1.1.3.5.; Colombie et al., 2008; Carmena et al., 2012). Dysfunction of Incenp in oocytes results in a drastic delay of spindle assembly and causes formation of ectopic poles. CPC has been proposed to promote microtubule accumulation near chromosomes in Drosophila oocytes in a similar way to in Xenopus egg extracts (see 1.1.2.2.; Radford et al., 2012; Maresca et al., 2009, Tseng et al., 2010). However, the molecular mechanism of Incenp function in vivo is not known.

Read Also

1.3.2. Chromosome positioning within the spindle

1.3.2.1. Arrangement of chromosomes in meiosis

In male and female meiosis, before anaphase of the first round of meiotic division, two homologous chromosomes of each pair are joined together (see 1.). The linkage is provided by cohesion along chromosome arms and by chiasmata at the site of recombination between the two homologs (Petronczki et al., 2003). As a result of the linkage a unique meiosis I chromosomal structure forms, called the bivalent. In metaphase I the bivalents are arranged so that sister kinetochores belonging to each homolog face and are attached to the same pole, while homologous chromosomes have bipolar attachments. This is opposite to amphitelic sister chromosome arrangement in meiosis II and mitosis (see 1. and 1.1.3.4.).

The arrangement of chromosomes in meiosis I remains stable until anaphase I (Hauf and Watanabe, 2004; Marston and Amon, 2004). In anaphase I, cohesion between chromosome arms is removed, however, cohesion between the sister centromeres remains. This allows homologous chromosomes separation from each other while the sister chromosomes remain joined at the centromere. In this way, sister chromosomes move towards the same spindle pole in anaphase. Cohesion between sister centromeres is resolved during anaphase II.

Paliulis and Nicklas (2000) showed that homologous chromosomes taken from a grasshopper spermatocyte in meiosis I and introduced into meiosis II cell segregated as in meiosis I. This indicated that the mode of meiotic chromosome segregation in the first round of division relies on the intrinsic features of the meiosis I chromosomes and not on the state cell cycle or on spindle organization.

Cohesin complex is known to link the sister centromeres (Hauf and Watanabe, 2004; Marston and Amon, 2004; Sakuno et al., 2009). It is composed of four conserved core subunits: Smc1, Smc3, Rec8 and Scc3 (Nasmyth and Haering, 2009). Cohesin complex forms a ring structure around sister DNA that physically joints them. Cohesin at sister centomeres persists until anaphase II owing to a conserved protein, MEI-S332/Shugoshin or by LAB-1 in holocentroc C. elegans (Kitajima et al., 2004; Watanabe, 2005; de Carvalho et al., 2008). Additionally, the metaphase I sister kinetochores have been shown by electron microscopy to be positioned side-by side (Goldstein, 1981; Parra et al., 2004). In budding yeast, this positioning is dependent on the monopolin complex, consisting of Mam1, Csm1 and Lsr4 subunits (Toth et al., 2000; Rabitsch et al., 2003). In S.pombe, the side-by-side orientation of sisters is partially regulated by a cohesin subunit, Rec8, and Moa1 (Watanabe and Nurse, 1999; Yokobayashi and Watanabe, 2005; Sakuno, et al, 2009). However, in higher eukaryotes the factors responsible for this side-by-side sister chromosomes arrangement have not been identified yet.

Despite the physical link of sister kinetochors in meiosis I, bipolar attachments of sister kinetochores still can be produced leading to missegregation of chromosomes (Hauf et al., 2007; Monje-Casas et al., 2007; Hassold and Hunt, 2001). The bipolarity of the fused sister kinetochores attachment is often a result of merotely, which in yeast meiosis I seems to be corrected in Aurora B-dependent fashion (see 1.1.3.5.; Hauf et al., 2007; Monje-Casas et al., 2007). In Drosophila oocytes, mutation of an Aurora B regulator, Incenp, results in frequent mono-orientation of bivalents (Resnick et al., 2009). It is however, unclear how Aurora B specifically stabilizes bipolar kinetochore-microtubule attachments of bivalents rather than sisters.

Based on studies in fission yeast, Sakuno et al. (2011) proposed a model that chiasmata promote geometry of the kinetochore, separating kinetochore-microtubule attachment site of the sisters from Aurora B-active region. According to this model, syntelic attachments of sisters are promoted by biorientation of homologs. Tension between bivalents is another possibility for stabilization of kinetochore-microtubule attachments of bivalents. In a study on grasshopper spermatocytes, the monopolar attachments of bivalents have been shown to be unstable and result in reorientation, followed by a bipolar attachment (Nicklas, 1997). However, pulling of a bivalent with monopolar attachment by micromanipulation, towards opposite pole to generate tension, appeared to stabilize the faulty attachment. These studies suggest that chiasmata may provide a physical restriction to the geometry of a bivalent. Additionally, as chiasmata join the chromosomes, their loss in Drosophila results in migration of homologs towards the poles (McKim et al., 1993; Jang et al., 1995).

Another level of complexity arises from the fact that achiasmatic homologous chromosomes, like the small fourth chromosome in Drosophila oocytes or balancer chromosomes, used routinely in Drosophila studies, arrange on the metaphase plate and segregate correctly. Drosophila male meiosis also proceeds without recombination and relies on other mechanisms of homologs conjunction than chiasmata (Thomas et al., 2005). Therefore, geometrical and physical restrictions provided by chiasmata may be redundant with other mechanisms involving Aurora B.

1.3.2.2. The role of kinetochores in female meiosis

In mitosis, the kinetochore is essential for capturing the spindle microtubules emanating from opposite poles to stabilize the spindle microtubule and to position chromosomes within the spindle (see 1.1.3.3.). However, the precise role of kinetochores in acentrosomal meiosis is not well defined. It is not clear to what degree kinetochores contribute to chromosome positioning within the spindle when spindle microtubule nucleation is mediated by chromosomes and when the spindle polarity is not pre-defined. Evidence from the past experiments indicates that kinetochores may not be as important for chromosome positioning in oocytes as they are in mitosis.

Electron microscopy in mouse oocytes demonstrated that spindle assembly and congression of chromosomes are achieved in the absence of kinetochore-fibers (see 1.3.1., Brunet et al., 1999). The full end-on attachment of kinetochores to microtubules occurs just before anaphase onset, suggesting that attachment of microtubules to kinetochores determines the anaphase onset. Live imaging in prometaphase mouse oocytes showed that already congressed chromosomes contact microtubules via kinetochores, however, these interactions are very erroneous and unstable (Kitajima et al., 2011). The nature and reason of the late establishment of kinetochore-microtubule stable interaction is not understood. Kitajima et al. (2011) propose that kinetochore-microtubule attachments are suppressed in order to prevent incorrect attachments during the complex process of bipolar spindle assembly. Nevertheless, it cannot be excluded that unstable lateral kinetochore-microtubule attachments may play a role in chromosome positioning before establishing stable end-on kinetochore-microtubule attachments on the metaphase plate (see 1.1.3.3.; Brunet et al., 1999).

Live imaging analysis revealed that in C. elegans female meiosis, kinetochores are dispensable for chromosome segregation but are crucial for arranging the chromosomes at the right angle relative to the spindle axis (Dumont et al., 2010). Nevertheless, chromosomes are congressed by kinetochore-independent mechanisms in late metaphase. It is noteworthy that C. elegans chromosomes are holocentric and kinetochores are not restricted to a single locus, but are spread along chromosome arms (see 1.1.3.1.; Maddox et al., 2004). Additionally, homologue separation seems to be mediated by extension of microtubules between the separating homologues (Dumont et al. 2010). Thus holocentric organisms may have different requirements for chromosome positioning and segregation than monocentric systems (Dumont et al., 2010).

Taken together, a kinetochore role in female meiosis is not well defined and may be different from mitotic. The mitotic role of a kinetochore may be compensated by kinetochore-independent mechanisms in female meiosis. These are likely to be provided by chromosome arms.

1.3.2.3. The role of kinetochore-independent interactions in female meiosis

In addition to the force exerted on chromosomes by kinetochore-microtubule attachment, kinetochore-independent force acting on chromosomes has been proposed to influence chromosome positioning within the spindle (see 1.1.3.3.; Rieder et al., 1986). Chromosome arms, separated from the kinetochore by laser microsurgery, were shown to move away from the pole. The force is known as the polar ejection force, PEF (Rieder and Salmon, 1994). Apart from the role in chromosome congression, PEF opposing kinetochore pole-ward pulling, can generate tension and stabilize microtubule attachment to kinetochore (see 1.1.3.4. and 1.1.3.5.; Cassimeris et al., 1994; Cane et al., 2013).

One suggested source of PEF is microtubule polymerization, which collides with chromosomes (Rieder and Salmon, 1994; McIntosh et al., 2002). In mitosis, microtubule density is highest at the poles. Therefore, it has been proposed that chromosomes are pushed away from the poles and locate in the spindle area with the lowest PEF, which is the spindle equator (Rieder and Salmon, 1994). However, in acentrosomal meiosis lacking defined MTOCs, like in Drosophila, microtubule density is the lowest at the poles (see 1.2.1.; Thekrauf and Hawley, 1992). This suggests that PEF distribution in female meiosis may differ greatly from in mitosis. Another possibility is that PEF may be biased towards spindle poles by centrosome-independent mechanism. In Drosophila oocytes, differently from in mitosis, the Augmin complex localizes to the spindle poles where it has been suggested to generate new microtubules (see 1.1.2.3.; Maireles et al., 2009). Deletion of an Augmin component, Wac, results in uncongressed chromosomes and low frequency of maloriented chromosomes. However, opposite to mitosis, in the wac mutant the spindle is robustly assembled (Goshima et al., 2008; Maireles et al., 2009). This suggests that in Drosophila, Augmin contributes to generation of PEF at the poles, specifically in oocytes. Additionally, unlike in mitosis, the Augmin function is essential for female meiosis, as the wac mutant leads to female sterility (see .1.2.3.; Maireles et al., 2009). Therefore, PEF may differently contribute to chromosome positioning in mitosis and meiosis.

Another complementary origin of PEF proposed is based on chromosome-microtubule interaction mediated by specialized chromosome-associated kinesins, called chromokinesins (Mazumdar and Misteli, 2005). Chromokinesins move chromosome arms towards or locate them at the pus ends of spindle microtubules, which means away from the poles. Disruption of chromokinesin function often affects chromosome arm congression in different systems. In mitotic cells and in Xenopus egg extract, several chromokinesins have been identified for their role in generation of PEF. Kinesin-10 family member, KID, is a motor protein proposed to walk chromosome arms along microtubules toward microtubule plus ends (Brouhard and Hunt, 2005). It was found to push chromosomes away from the poles in Xenopus egg extract (Antonio et al., 2000; Funabiki and Murray, 2000). In human culture cells, KID has an effect on the position of chromosome arms but not on kinetochore position and it is not required for chromosome segregation (Levesque and Compton, 2001). A non-motile kinesin-10 family member, Nod, has been proposed to participate in chromosome congression by end-tracking of polymerizing microtubules in Drosophila mitotic cells (Cochran et al., 2009). Depletion of Nod results in chromosome arms extended from equator toward spindle poles (Goshima and Vale, 2003). Kinesin-4 family member, KLP-19 is a motor required for chromosome congression in C.elegans (Powers et al., 2004).

Studies on mitotic cells and in Xenopus egg extracts do not provide unambiguous evidence that chromokinesins are essential in monocentric organisms, and so their biological objective in mitotic cells, other than in C.elegans, is not clear. However, several studies indicate that they may be particularly important in acentrosomal female meiosis. In mouse oocytes, kinetochores do not seem to play a crucial role in chromosome congression (see 1.3.2.2.; Brunet et al., 1999). Instead, chromatin-microtubule interactions locate chromosomes on the spindle equator (Kitajima et al., 2011). Based on live imaging of chromosome behaviour, Kitajima et al. (2011) suggests that chromokinesins play an important role in chromosome individualisation in the early stages of spindle assembly. However, the exact mechanism and chromokinesins responsible for this process are not known. KID has proved itself dispensable for chromosome positioning in mouse oocytes (Kitajima et al., 2011).

Like in mitosis, in Drosophila oocytes, Nod has been proposed to produce PEF influencing chromosome congression (Therkauf and Hawley, 1992; Matthies et al., 1999). Nod-generated PEF is crucial for congression of achiasmatic chromosome, but has no obvious effect on chiasmatic chromosomes (Therkauf and Hawley, 1992; Matthies et al., 1999).

In C.elegans oocytes, thick microtubule bundles along chromosomes have been proposed to play a role in biorientation and congression of chromosomes (Wignall and Villeneuve, 2009). Together with the fact that kinetochores do not seem required for these processes, a role of chromokinesins in chromosome positioning is a real possibility. Concomitantly, Wignall and Villeneuve (2009) recorded severe chromosome congression defects after depletion of the chromokinesin Klp19. This drastic effect was, however, not observed by Dumont et al. (2010).

Despite the fact that chromosome arm-microtubule interactions may be particularly important for chromosome positioning in acentrosomal spindles, the origin of PEF and the extent of PEF contribution to this process remain unclear.