BIOLOGICAL OVERVIEW

Accurate segregation of chromosomes is critical to ensure that each daughter cell receives the full genetic complement. Maintenance of cohesion between sister chromatids, especially at centromeres, is required to segregate chromosomes precisely during mitosis and meiosis. The Drosophila protein Mei-S332, the founding member of a conserved protein family, is essential in meiosis for maintaining cohesion at centromeres until sister chromatids separate at the metaphase II/anaphase II transition. Mei-S332 localizes onto centromeres in prometaphase of mitosis or meiosis I, remaining until sister chromatids segregate. A mechanism has been elucidated for controlling release of Mei-S332 from centromeres via phosphorylation by Polo kinase. Polo antagonizes Mei-S332 cohesive function and full Polo activity is needed to remove Mei-S332 from centromeres, yet this delocalization is not required for sister chromatid separation. Polo phosphorylates Mei-S332 in vitro, Polo and Mei-S332 bind each other, and mutation of Polo binding sites prevents Mei-S332 dissociation from centromeres (Clarke, 2005).

The accurate segregation of chromosomes is essential to prevent aneuploidy following cell division. Cohesion between sister chromatid centromeres is necessary to ensure that the two sister chromatid kinetochores attach stably to microtubules emanating from opposite spindle poles. In mitosis, cohesion at centromeres and along the chromosome arms is maintained between sisters until the metaphase/anaphase transition. In meiosis, chromosome arm cohesion plays an additional role in linking homologs together by stabilizing chiasmata. Arm cohesion is dissolved at the metaphase I/anaphase I transition, permitting homolog segregation, and only centromere cohesion remains between sisters until the metaphase II/anaphase II transition when the sisters finally separate. Therefore, the maintenance of centromere cohesion is essential to prevent chromosome missegregation in meiosis, yet the molecular mechanisms underlying the control of this cohesive force are poorly understood (Clarke, 2005).

The cohesin complex is required for sister chromatid cohesion. This complex is composed of four subunits, Smc1, Smc3, Scc1 (Rad21), and Scc3, and in meiosis Rec8 replaces the Scc1 subunit. In some organisms, other additional meiotic-specific subunits replace those employed in mitosis. These proteins are thought to form a ring structure that loops around the two sisters and holds them together. At the metaphase/anaphase transition in mitosis, the cysteine protease separase (see Drosophila Separase) is activated via the anaphase promoting complex/cyclosome (APC/C)-dependent destruction of its inhibitor securin. It then cleaves Scc1 to release cohesion between sisters. In yeast during meiosis, separase becomes activated at the metaphase I/anaphase I transition and cleaves arm cohesin. A portion of the cohesin complex remains uncleaved at centromeres, keeping sister chromatids together until the metaphase II/anaphase II transition, when this persistent cohesin is cleaved (Clarke, 2005 and references therein).

Although it is likely that protector proteins prevent cleavage of cohesin at centromeres until the metaphase II/anaphase II transition, their identity remains elusive. The best candidate is the Drosophila centromere cohesion protein Mei-S332 because of its striking mutant phenotype and localization pattern. Mutations in mei-S332 lead to loss of sister chromatid cohesion at the centromere beginning at anaphase I, resulting in chromosome nondisjunction during meiosis II (Davis, 1971; Goldstein, 1980 and Kerrebrock, 1992). Consistent with a direct role in regulating cohesion, Mei-S332 localizes to meiotic centromeres from prometaphase I to metaphase II, dissociating concomitantly with segregation of sister chromatids (Kerrebrock, 1995 and Moore, 1998). Mei-S332 also localizes to mitotic chromosomes from prometaphase to metaphase and plays a modest role in sister chromatid cohesion during mitosis (LeBlanc, 1999; Moore, 1998; Clarke, 2005 and references therein).

Until recently, Mei-S332 was thought to be a unique Drosophila centromere protein. It now has been recognized as the founding member of a newly identified family of proteins conserved from yeast to humans. The Mei-S332 homolog Sgo1 was identified in budding and fission yeast by its ability to protect centromere cohesion, its expression during meiosis, and its localization to centromeres in meiosis I (Katis, 2004; Kitajima, 2004; Marston; 2004 and Rabitsch et al. 2004). Based on sequence identity/similarity, the Sgo1 protein is present in many other organisms including humans, and a related protein, Sgo2, has been identified in S. pombe and Arabidopsis. Sgo1 has been proposed to protect the centromeric meiosis-specific subunit Rec8 from separase cleavage at the metaphase I/anaphase I transition. In contrast, in fission yeast Sgo2 promotes mono-orientation of sister chromatid kinetochores to ensure that the two sister chromatids of each homolog migrate to the same pole at anaphase I (Rabitsch, 2004). Sgo1 affects mitotic segregation in budding yeast, and Sgo2 functions in mitosis in fission yeast (Katis, 2004; Kitajima, 2004 and Marston, 2004). In vertebrate cells, Sgo1 is needed to maintain centromere cohesion in mitosis and also affects spindle microtubule dynamics (Salic, 2004; Clarke, 2005 and references therein).

A candidate regulator for the Mei-S332 family is the highly conserved Polo kinase. Polo controls many aspects of mitosis, such as the onset of mitosis, spindle formation, the metaphase/anaphase transition, and cytokinesis. One crucial Polo substrate is the Scc1/Rad21 cohesin subunit, which accounts for the metaphase arrest and failure to release sister chromatid cohesion seen in polo mutants. Phosphorylation of Rad21 by the Xenopus homolog of Polo, Plx1, is responsible for the separase-independent removal of the bulk of cohesin along sister chromatid arms in prophase. In budding yeast, phosphorylation of Scc1 by the Polo ortholog Cdc5 near separase cleavage sites enhances its cleavage by separase (Clarke, 2005 and references therein).

Polo also is critical for unique aspects of meiosis. Cdc5 is required for complete phosphorylation and subsequent cleavage of Rec8. It also regulates kinetochore orientation by phosphorylating Mam1, a meiosis-specific kinetochore protein that is part of a protein complex needed for sister kinetochores to mono-orient toward the same spindle pole in prometaphase I. In cdc5 mutants Mam1 fails to localize to kinetochores. In addition, Cdc5 may be necessary for the resolution of recombination intermediates into crossovers prior to meiosis I (Clarke, 2005 and references therein).

To understand the mechanism that controls cohesion at centromeres, how Mei-S332 function and localization are regulated in mitosis and meiosis were studied. It was found that Polo kinase plays an essential role in Mei-S332 dissociation from centromeres (Clarke, 2005).

The results demonstrate that Polo kinase regulates Mei-S332 localization and aspects of its function. polo mutants dominantly suppress the mei-S332⁸ nondisjunction phenotype and wild-type Mei-S332 is retained at centromeres past the metaphase II/anaphase II transition in these polo mutants. Mei-S332 appears to be phosphorylated in mitosis at the metaphase/anaphase transition, and in polo mutants, Mei-S332 persists on centromeres into interphase, consistent with phosphorylation being a signal for Mei-S332 to delocalize. Polo kinase binds to Mei-S332, and this is partly dependent on two PBD binding site motifs. Furthermore, in vitro phosphorylation of Mei-S332 is dependent on at least one motif and dependent on Polo. These two PBD binding site motifs in vivo are likely required for chromosomal dissociation of Mei-S332, similar to the effects observed in polo mutants. Together, these data point to Polo as a key regulator of Mei-S332 centromere localization in both mitosis and meiosis (Clarke, 2005).

Polo function is required for Mei-S332 delocalization from centromeres in mitosis and meiosis, but the ability of polo mutants to dominantly suppress mei-S332 mutants additionally shows that Polo antagonizes Mei-S332 function. This is important because the results indicate that cohesion can be released even if Mei-S332 remains localized in polo mutants. Mei-S332 remains on centromeres after the metaphase II/anaphase II transition in polo/+ mutants, yet reasonably normal disjunction of chromosomes occurs during meiosis II in these heterozygotes. In mitosis, Mei-S332 can remain on the centromeres of the fourth chromosomes in polo mutants even when sister chromatid cohesion is released and they segregate to the poles. Thus, Polo phosphorylation is necessary for delocalization of Mei-S332, but there must exist a mechanism to inactivate Mei-S332 to release cohesion that is independent of Mei-S332 dissociation. This pathway is not entirely dependent on Polo, although Polo may contribute (Clarke, 2005).

The idea that Mei-S332 can remain localized to centromeres without cohesion between sister chromatids is supported by several examples. In double parked mutants, Mei-S332 localizes to unreplicated, single chromatids on which cohesion has never been established (Lee, 2004). This shows that the presence of sister chromatid cohesion is not a prerequisite for Mei-S332 localization to centromeres. Similarly, Mei-S332 localizes to single sister chromatids in ord mutants, in which sister chromatids separate prematurely early in meiosis I (Bickel, 1998). Finally, Sgo1 can localize to centromeres in early anaphase II when the 3′ UTR of Sgo1 is disrupted, yet no interference of the release of sister chromatid cohesion is observed (Rabitsch, 2004). This result suggests that Sgo1 can promote sister chromatid cohesion in meiosis and can subsequently be inactivated yet remain at centromeres (Clarke, 2005).

It is proposed that Mei-S332 centromere localization is regulated by phosphorylation. In this model, the assembly of Mei-S332 onto centromeres in prometaphase I is controlled by the action of an unknown phosphatase. Mei-S332 remains localized to centromeres until the metaphase II/anaphase II transition when Polo kinase binds to Mei-S332, via the phosphorylated T331 PBD binding site, and phosphorylates Mei-S332 elsewhere, initiating Mei-S332 dissociation from centromeres. These data suggest that both S234 and T331 contribute to Polo binding and to Mei-S332 centromere dissociation, but in vitro T331 plays the predominant role in Plx1-dependent phosphorylation. Further, it is suggested that Polo functions to antagonize Mei-S332 activity in meiosis, thereby affecting the release of sister chromatid cohesion. This may be either through phosphorylation of Mei-S332 or by affecting another component of sister chromatid segregation. Based on the results, it cannot be distinguished whether phosphorylation of Mei-S332 by Polo antagonizes Mei-S332 activity directly (Clarke, 2005).

It is proposed that Polo directly phosphorylates Mei-S332 because the proteins can bind each other. Importantly, this binding is reduced by disruption of the PBD binding site motifs, and these mutations abolish Plx1-dependent phosphorylation of Mei-S332 and prevent Mei-S332 from dissociating from centromeres in S2 cells. PBD binding sites are required to be phosphorylated in order for Polo to bind to its substrates. Given that Plx1-dependent phosphorylation of Mei-S332 was disrupted by mutating this site, it would seem that T331 and S234 need to be phosphorylated prior to Polo binding to Mei-S332, and then subsequent unknown sites on Mei-S332 can be phosphorylated by Polo kinase, thereby dissociating Mei-S332 from centromeres. In support of this idea, it has recently been proposed that once Polo binds to a substrate via an interaction with a PBD binding site, its kinase activity toward that substrate is stimulated (Clarke, 2005).

Which kinase is responsible for phosphorylating S234/T331 initially? Candidate kinases are a cyclin-dependent kinase, with specificity for sites with a proline in the +1 position as at the T331 and S234 sites, or a kinase such as Aurora B, which is localized to centromeres at the metaphase/anaphase transition in mitosis. The control of this precise phosphorylation event would effectively prevent Polo from phosphorylating Mei-S332 and releasing it from centromeres until the appropriate time. Correlating with this possibility, during both mitosis and meiosis, Polo is poised at centromeres yet does not act to remove Mei-S332 until the proper time. In mitosis, Mei-S332 becomes localized to centromeres in prometaphase (Moore, 1998) and Polo kinase is also localized to centromeres at this time. In meiosis, from metaphase I to metaphase II Polo localizes to centrosomes and centromeres, yet Mei-S332 is not released until metaphase II/anaphase II (Kerrebrock, 1995; Clarke, 2005).

Mei-S332 is the founding member of a family of proteins required for maintaining centromere cohesion between sister chromatids. This study has defined Polo kinase as crucial for delocalization of Mei-S332. Polo antagonizes Mei-S332 activity. These results strongly indicate that Polo directly phosphorylates Mei-S332 and that this leads to delocalization. It will be interesting to identify the anchor for Mei-S332 centromere binding and to decipher how phosphorylation of Mei-S332 affects this interaction. These results additionally uncover a mechanism distinct from delocalization to inactivate Mei-S332 (Clarke, 2005).

GENE STRUCTURE

cDNA clone length - 1540

Bases in 5' UTR - 213

Exons - 3

Bases in 3' UTR - 121

PROTEIN STRUCTURE

Amino Acids - 401

Structural Domains

For information on Mei-S223 structure see Control of centromere localization of the Mei-S332 cohesion protection protein

date revised: 29 May 2005

Home page: The Interactive Fly © 1995, 1996 Thomas B. Brody, Ph.D.

The Interactive Fly resides on the
Society for Developmental Biology's Web server.