Meiotic regulation of the Ndc80 complex composition and function

This review describes the current models for how the subunit abundance of the Ndc80 complex, a key kinetochore component, is regulated in budding yeast and metazoan meiosis. The past decades of kinetochore research have established the Ndc80 complex to be a key microtubule interactor and a central hub for regulating chromosome segregation. Recent studies further demonstrate that Ndc80 is the limiting kinetochore subunit that dictates the timing of kinetochore activation in budding yeast meiosis. Here, we discuss the molecular circuits that regulate Ndc80 protein synthesis and degradation in budding yeast meiosis and compare the findings with those from metazoans. We envision the regulatory principles discovered in budding yeast to be conserved in metazoans, thereby providing guidance into future investigations on kinetochore regulation in human health and disease.


Introduction
The kinetochore is an evolutionarily conserved multi-subunit protein complex that mediates chromosome segregation. Since the discovery of the kinetochore in 1980s, this large protein complex has emerged as a crucial regulatory hub that directs faithful genome partitioning. Its mis-regulation now serves as a biomarker for cancer. The past 30 years of kinetochore research have illuminated the structure and function of the kinetochore. Until recently, however, relatively little is known about how the abundance of each kinetochore subunit is regulated in specialized cellular contexts, such as meiosis. In this review, we focus on the regulation of the protein abundance of one kinetochore subcomplex, the Ndc80 complex, which dictates when kinetochores are active in meiosis.
In a simplified view, kinetochore subunits are organized into two main parts: the inner and outer kinetochore. The inner kinetochore forms at the centromere, providing the foundation for the outer kinetochore to assemble. The outer kinetochore interacts with dynamic spindle microtubules.
Given the importance of Ndc80, it is perhaps not surprising that Ndc80 has evolved to become the linchpin subunit of the outer kinetochore in budding yeast meiosis, the specialized cell division that generates gametes. In normal meiosis, one round of DNA replication is followed by two consecutive chromosome divisions: homologous chromosomes are segregated in meiosis I, and then sister chromatids are pulled apart in meiosis II. In budding yeast, all subunits of the Ndc80 complex, with the exception of Ndc80, have constant protein levels throughout meiosis (Meyer et al. 2015;Chen et al. 2017). In meiotic prophase, Ndc80 protein levels decline, which results in the disassembly of the outer kinetochore. Shortly before the first meiotic division, Ndc80 levels increase to allow outer kinetochore re-assembly, just in time for chromosome segregation (Asakawa et al. 2005;Miller et al. 2012;Meyer et al. 2015;Chen et al. 2017Chen et al. , 2020. This dynamic behavior of the outer kinetochore allows two key events to take place in meiosis I. First, the kinetochores of the sister chromatid pair attach to the spindle microtubules emanating from the same spindle pole (monopolar attachment) (Meyer et al. 2018). Second, the centromeric cohesion between sister chromatids are protected during anaphase I (Miller et al. 2012). Premature expression of Ndc80 in meiotic prophase can alter chromosome segregation pattern such that sister chromatids segregate in meiosis I. This abnormal meiosis leads to both defective and reduced number of gametes (Miller et al. 2012). Therefore, a key aspect of establishing meiosis I boils down to regulating Ndc80 protein levels, which are intricately controlled by the synthesis and degradation of Ndc80 proteins in meiosis.

Ndc80 synthesis: toggling of two functionally distinct mRNAs
To dial up or down Ndc80 abundance, meiotic yeast cells modulate Ndc80 synthesis by controlling the levels of two mRNA isoforms expressed from the NDC80 gene Chia et al. 2017). These two mRNA isoforms share the entire coding sequence of NDC80 but vary in their 5′ end. The longer isoform, named the long undecoded transcript isoform (NDC80 LUTI ), has a 5′-extension that contains nine upstream open reading frames (uORFs). Translation of the uORFs prevents ribosomes from accessing the main ORF; consequently, NDC80 LUTI cannot be translated into Ndc80 protein. The shorter, canonical mRNA (NDC80 ORF ) lacks the 5′-extension and is capable of Ndc80 protein production. Rather than being a protein-coding unit, NDC80 LUTI serves a regulatory function such that its transcription inactivates the canonical NDC80 promoter through co-transcriptional histone marks and nucleosome re-positioning. As a result, upregulation of NDC80 LUTI expression causes downregulation of Ndc80 protein synthesis and hence, downregulation of the NDC80 gene.
Meiotic cells control the relative expression of two NDC80 mRNA isoforms through the action of two key meiotic transcription factors. Budding yeast meiosis is induced by nutrient deprivation. Upon meiotic entry, the transcription factor Ime1 is upregulated and binds to Ume6 (Kassir et al. 1988;van Werven and Amon 2011). This Ime1-Ume6 transcription factor complex drives expression of NDC80 LUTI and early meiotic genes (Fig. 1). NDC80 LUTI transcription in turn shuts down NDC80 ORF , which is expressed before meiosis onset, and hence inhibits Ndc80 protein synthesis. Meanwhile, Ndc80 turnover is upregulated in meiotic prophase (described in the next section). Due to the dual action of synthesis repression and degradation enhancement, the protein level of Ndc80 precipitously drops in meiotic prophase, leading to outer kinetochore disassembly. After cells exit from meiotic prophase, the mid-meiotic transcription factor Ndt80 (Xu et al. 1995;Chu and Herskowitz 1998) induces NDC80 ORF expression. Since NDC80 ORF is capable of translating Ndc80 protein, Ndc80 is rapidly resynthesized to allow assembly of the Ndc80 complex onto the inner kinetochore and reactivation of the kinetochore for mediating meiotic chromosome segregation. Therefore, the timely fluctuation of Ndc80 levels (hence outer kinetochore assembly) is ensured by coupling the expression of NDC80 LUTI and NDC80 ORF to master meiotic transcription factors.
Several mechanistic details of the NDC80 LUTI -based repression remain unresolved. For example, nucleosomes are repositioned at the canonical NDC80 ORF promoter in response to NDC80 LUTI expression; however, the identity of the chromatin remodeler(s) responsible for this change is unknown. In addition, it is not understood how the NDC80 LUTI -based repression mechanism is turned off after NDC80 LUTI expression subsides at prophase exit, the same time when the transcription factor Ndt80 re-induces NDC80 ORF . Do the histone modifications and nucleosome positioning reset to the pre-meiotic state? If so, does the transcription factor Ndt80, or histone demethylase(s), or nucleosome remodeler(s) play a role in resetting the chromatin states? Deeper mechanistic understanding into how LUTI-based repression is established and erased will help generate models to predict LUTIs and uncover LUTI regulators in other genomes.

Ndc80 degradation: a new function for the conserved kinase-substrate pair
In concert with the repression of Ndc80 synthesis in meiotic prophase, the premeiotic pool of Ndc80 proteins is also degraded in a regulated manner (Chen et al. 2020). Central to Ndc80 degradation is the phosphorylation of Ndc80 by Aurora B/Ipl1 (Fig. 1). Aurora B/Ipl1 is the catalytic subunit of the chromosome passenger complex, which regulates 1 3 diverse cellular processes, including the assembly of bipolar spindles and kinetochore, kinetochore orientation, and cytokinesis (reviewed in Lampson and Grishchuk 2017). The functional significance of Ndc80 phosphorylation has been primarily characterized in mitotic cells: Aurora B-dependent phosphorylation weakens Ndc80's binding to microtubules, leading to kinetochore detachment and an opportunity for the kinetochore to re-attach to microtubules in the correct orientation (reviewed in Biggins 2013; Wimbish and DeLuca 2020).
In budding yeast meiotic prophase, phosphorylated Ndc80 is a target for degradation. The Aurora B/Ipl1dependent Ndc80 phosphorylation triggers a set of downstream events that depend on a short N-terminal segment of Ndc80 (the 2-28 residues) and the meiotic ubiquitin ligase APC Ama1 , ultimately driving Ndc80 for proteasomal degradation (Chen et al. 2020). Notably, Ndc80 phosphorylation results in kinetochore detachment from microtubules in meiotic prophase, similar to its role in mitosis. And yet, it is unlikely that this detachment directly drives Ndc80 degradation in meiotic prophase, as microtubule depolymerization cannot rescue the defect in Ndc80 degradation resulting from Aurora B/Ipl1 depletion (Chen et al. 2020).
Besides regulating Ndc80 degradation, Aurora B/Ipl1 also plays an important role in suppressing kinetochore-microtubule interactions in meiotic prophase. Aurora B/Ipl1 does so by preventing bipolar spindle formation in meiotic prophase (Shirk et al. 2011;Kim et al. 2013). During this meiotic stage, Aurora B/Ipl1 localizes to the spindle pole bodies (SPBs, the centrosomes in yeast) and the short nuclear microtubule arrays. It has been proposed that these Aurora B/Ipl1 molecules prevent premature separation of the duplicated SPBs and formation of bipolar spindles (Shirk et al. 2011;Kim et al. 2013 activity and irreversibly abolishing the microtubule-binding site of the kinetochore through Ndc80 degradation, Aurora B/Ipl1 ensures that the kinetochore interacts with spindle microtubules only after meiotic prophase. This delayed kinetochore-microtubule interaction is required to set up a meiosis I-specific chromosome segregation pattern (Miller et al. 2012). Several mechanistic details of Ndc80 degradation remain unresolved. For example, it is unknown how the 2-28 residues of Ndc80 mediate Ndc80 degradation. These 27 residues are necessary for Ndc80 degradation and not for Aurora B/Ipl1-dependent phosphorylation. Importantly, this region, together with the Aurora B/Ipl1 phosphorylation sites, is sufficient to induce the degradation of another kinetochore protein (Chen et al. 2020). We posit that Ndc80 phosphorylation may recruit a bipartite protein or a protein complex that binds to the 2-28 region, or Ndc80 phosphorylation may induce local conformational changes to expose the 2-28 region for factor binding. Alternatively, the 2-28 region may act in parallel with Ndc80 phosphorylation to signal for Ndc80 degradation.
Another fascinating question is which pool of Ndc80 protein is most efficiently targeted by the degradation pathway: soluble Ndc80 or the Ndc80 protein that is part of the Ndc80 complex (on or off the kinetochore)? In one model, the degradation pathway only acts on the soluble pool of Ndc80. Molecular factors would first extract Ndc80 from its complex, and Ndc80 gets degraded once it becomes soluble. In another model, Ndc80 is locally degraded at the kinetochore, while soluble Ndc80 is protected from degradation. It would be interesting to directly measure the degradation rate of soluble Ndc80 proteins versus those localized to the kinetochore to determine whether Ndc80's kinetochore localization is required for its degradation.
Also unknown is the role of APC Ama1 . While APC Ama1 is a meiosis-specific ubiquitin ligase, whether Ndc80 is a direct substrate of APC Ama1 remains to be tested. If Ndc80 is a direct substrate of APC Ama1 , it would be interesting to test how Aurora B/Ipl1-dependent phosphorylation of Ndc80 affects the interaction between Ndc80 and APC Ama1 .
Lastly, Ndc80 degradation is turned off in metaphase I through an unknown mechanism (Chen et al. 2020). We propose two non-mutually exclusive models. First, the completion of error correction may repress Ndc80 degradation. After the chromosomes correctly attach to the spindle microtubules, phosphatases remove Ndc80 phosphorylation to stabilize the microtubule attachments (reviewed in Biggins 2013). Consequently, the trigger for Ndc80 degradation (phosphorylation) would be eliminated. Second, Clb-CDK activity becomes elevated in prometaphase I, leading to inactivation of the ubiquitin ligase APC Ama1 (Oelschlaegel et al. 2005;Tsuchiya et al. 2011). Without APC Ama1 , Ndc80 would be stabilized even before the completion of error correction. Future studies will be necessary to test these models.

Regulation of the Ndc80 complex in metazoan meiosis
The basic composition and many subunits of the kinetochore are conserved between yeast and metazoans (reviewed in Musacchio and Desai 2017). The inner kinetochore of metazoans consists of the Constitutive Centromere Associated Network (CCAN). Most of the CCAN subunits have orthologs in yeast along with additional metazoan specific proteins (reviewed in Musacchio and Desai 2017). CCAN binds to centromeric chromatin throughout the cell cycle. The metazoan outer kinetochore is called the KMN network (Knl1/Mis12/Ndc80 complexes). The protein structure and core domains of the Ndc80 complex are conserved between yeast and metazoans (Wei et al. 2007). Also conserved is the role of Aurora B in correcting the erroneous attachments between the Ndc80 complex and spindle microtubules (Cheeseman et al. 2006;DeLuca et al. 2006DeLuca et al. , 2011Alushin et al. 2012;Zaytsev et al. 2014Zaytsev et al. , 2015. One difference, however, lies in the assembly timing of the Ndc80 complex during cell division: the Ndc80 complex assembles onto CCAN only after nuclear envelope breakdown in metazoans (Gascoigne and Cheeseman 2013) while the yeast Ndc80 complex associates with the inner kinetochore during most of the mitotic cell cycle except briefly in S phase (Kitamura et al. 2007). In HeLa cells, the assembly and disassembly of the Ndc80 complex are controlled by at least two means. First, the Ndc80 complex is excluded from the nucleus until after nuclear envelope breakdown (Gascoigne and Cheeseman 2013). Second, high CDK activity promotes the Ndc80 complex to assemble on the kinetochore by phosphorylating CENP-T, which enhances the binding between CENP-T and the Ndc80 complex (Gascoigne et al. 2011;Nishino et al. 2013;Huis In't Veld et al. 2016). At mitotic exit, the declined CDK activity dampens the interaction between CENP-T and Ndc80, resulting in reduced levels of the Ndc80 complex at the kinetochore. The timely disassembly of the Ndc80 complex is required for faithful chromosome segregation in subsequent cell divisions (Gascoigne and Cheeseman 2013).
In metazoan meiosis, the Ndc80 complex also disassembles during meiotic prophase as in yeast meiosis. For example, in mice oogenesis, Nuf2 and Spc24 form puncta on chromosomes (indicating kinetochore localization) only after germinal vesicle breakdown, which corresponds to an exit from prophase I (Zhang et al. 2015. The localization pattern of Ndc80 and Spc25 is less clear since antibody staining did not show distinct puncta on chromosomes even during metaphase I and metaphase II (Sun et al. 2010(Sun et al. , 2011. While there is no direct evidence currently, Ndc80 and Spc25 likely follow the localization pattern of Nuf2 and Spc24 because the four subunits form a complex both in vitro and in vivo (Janke et al. 2001;Wigge and Kilmartin 2001;Ciferri et al. 2005;Wei et al. 2005;Hornung et al. 2011). The mechanistic details are unknown for how mice oogenesis ensures the timely assembly of the Ndc80 complex. It would be interesting to explore how the synthesis, degradation, and nuclear localization of the Ndc80 complex are regulated during oogenesis in this organism.
In C. elegans oogenesis, the KMN subunits (Spc105/ KNL-1 and Nuf2/HIM-10) are first detected throughout chromatin in late pachytene/diplotene and transition to chromosome surface immediately before the oocytes enter spermatheca, the time of nuclear envelope breakdown in oocytes (Howe et al. 2001;Monen et al. 2005). In metaphase I, the KMN (NDC-80, Nuf2/HIM-10, MIS-12, KNL-1, and KNL-3) cups the poleward ends of chromosomes. Surprisingly, these proteins disappear from the chromosomes after the onset of anaphase I and reappear only after metaphase II (Monen et al. 2005;Dumont et al. 2010;Davis-Roca et al. 2017). Consequently, while the outer kinetochore facilitates response to erroneous microtubule-kinetochore attachments, it is not required for segregating chromosomes in the meiosis I of C. elegans oocytes (Dumont et al. 2010;Muscat et al. 2015;Davis-Roca et al. 2017;Laband et al. 2017).
How the KMN disassembles during anaphase I in C. elegans oogenesis remains unknown. It has been shown that this disassembly requires the Y-complex nucleoporin MEL-28, which recruits protein phosphatase 1 to the kinetochore (Hattersley et al. 2016). In addition, a protein complex known as the midbivalent ring may mediate KMN disassembly. As the KMN, the midbivalent rings are also removed from chromosomes in anaphase I (Dumont et al. 2010;Muscat et al. 2015), and defective ring removal correlates with delayed kinetochore disassembly (Davis-Roca et al. 2017). Interestingly, Aurora B/AIR-2 localizes to these rings (Wignall and Villeneuve 2009;Davis-Roca et al. 2017). Inspired by the finding in yeast meiosis, we propose that Aurora B might trigger kinetochore degradation to disassemble the outer kinetochore during anaphase I in C. elegans oocytes.
Beyond C. elegans and mice oogenesis, less is known about the timing and regulation of the Ndc80 complex assembly during meiosis in other organisms. Any differences during the meiosis of different sexes (male, female, hermaphrodites, etc.) are also not well characterized. Interestingly, the localization and function of some spindle assembly checkpoint proteins display sex-specific differences in mice. For example, Mad2 localizes to kinetochores throughout meiosis I in males but is lost from kinetochores in females (Kallio et al. 2000). While oocyte meiosis requires Bub3 for accurate chromosome segregation (Li et al. 2009), spermatocyte meiosis does not (Jeganathan and van Deursen 2006).
Given that the kinetochore is a key component of the spindle assembly checkpoint (reviewed in Musacchio and Desai 2017), it would be interesting to test whether the assembly and/or function of specific kinetochore subunits also differ between male and female meiosis.

Conclusions
As a central hub for ensuring accurate chromosome segregation, the kinetochore is subjected to intricate regulation of its subunit abundance, composition, assembly, and activity. Studies of budding yeast meiosis have revealed an integrated network that regulates the synthesis and degradation of one specific kinetochore subunit, Ndc80. Since Ndc80 is a limiting component of the yeast meiotic kinetochores, its timely synthesis and degradation determines when the kinetochores are active in meiosis. Whether similar regulation of Ndc80 abundance occurs in metazoan meiosis remains to be tested. Interestingly, proteins that localize to the kinetochores (e.g., spindle assembly checkpoint proteins) seem to have sexspecific dynamics and functions. It would be interesting to examine how regulation on kinetochore subunit abundance contributes to the observed differences in male and female meiosis.
More broadly, it is of great importance to understand the regulatory mechanisms of kinetochore abundance since altered levels of kinetochore proteins have been observed in many types of cancer. For example, for many kinetochore subunits, their depletion hinders timely chromosome segregation and causes lagging chromosomes at the cleavage furrow (reviewed in Ganem and Pellman 2012;Biggins 2013). Genome instability occurs when the inner kinetochore component CENP-A is overexpressed in flies and human cells (Heun et al. 2006;Au et al. 2008;Shrestha et al. 2017). The excess CENP-A proteins, along with additional kinetochore subunits, mislocalize to non-centromeric regions. These ectopic kinetochores put chromosome arms under tension due to spindle microtubule interactions and/or reduce the protein levels of kinetochore subunits at centromeres, leading to chromosome segregation errors. Additionally, overexpression of the outer kinetochore subunits Hec1/ Ndc80 or SKA1 is associated with multiple cancers and tumorigenesis (Chen et al. 1997;Chen et al. 2018;Hayama et al. 2006;Li et al. 2014;Shen et al. 2016). These studies highlight the significance of maintaining a proper level and stoichiometry of kinetochore subunits in ensuring accurate chromosome segregation. A better understanding of these regulatory pathways can provide new molecular targets for cancer treatments.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.