Mck1-mediated proteolysis of CENP-A prevents mislocalization of CENP-A for chromosomal stability in Saccharomyces cerevisiae.

Centromeric localization of evolutionarily conserved CENP-A (Cse4 in Saccharomyces cerevisiae) is essential for chromosomal stability. Mislocalization of overexpressed CENP-A to non-centromeric regions contributes to chromosomal instability (CIN) in yeasts, flies, and humans. Overexpression and mislocalization of CENP-A observed in many cancers is associated with poor prognosis. Previous studies have shown that F-box proteins, Cdc4 and Met30 of the Skp, Cullin, F-box (SCF) ubiquitin ligase cooperatively regulate proteolysis of Cse4 to prevent Cse4 mislocalization and CIN under normal physiological conditions. Mck1-mediated phosphorylation of SCF-Cdc4 substrates such as Cdc6 and Rcn1 enhances the interaction of the substrates with Cdc4. Here, we report that Mck1 interacts with Cse4, and Mck1-mediated proteolysis of Cse4 prevents Cse4 mislocalization for chromosomal stability. Our results showed that mck1Δ strain overexpressing CSE4 (GAL-CSE4) exhibits lethality, defects in ubiquitin-mediated proteolysis of Cse4, mislocalization of Cse4 and reduced Cse4-Cdc4 interaction. Strain expressing GAL-cse4-3A with mutations in three potential Mck1 phosphorylation consensus site (S10, S16, and T166) also exhibits growth defects, increased stability with mislocalization of Cse4-3A, CIN, and reduced interaction with Cdc4. Constitutive expression of histone H3 (Δ16H3) suppresses the CIN phenotype of GAL-cse4-3A strain, suggesting that the CIN phenotype is linked to Cse4-3A mislocalization. We conclude that Mck1 and its three potential phosphorylation sites on Cse4 promote Cse4-Cdc4 interaction and this contributes to ubiquitin-mediated proteolysis of Cse4 preventing its mislocalization and CIN. These studies advance our understanding of pathways that regulate cellular levels of CENP-A to prevent mislocalization of CENP-A in human cancers.

Structural basis for Mis18 complex assembly and its implications for centromere maintenance.

The centromere, defined by the enrichment of CENP-A (a Histone H3 variant) containing nucleosomes, is a specialised chromosomal locus that acts as a microtubule attachment site. To preserve centromere identity, CENP-A levels must be maintained through active CENP-A loading during the cell cycle. A central player mediating this process is the Mis18 complex (Mis18α, Mis18ß and Mis18BP1), which recruits the CENP-A-specific chaperone HJURP to centromeres for CENP-A deposition. Here, using a multi-pronged approach, we characterise the structure of the Mis18 complex and show that multiple hetero- and homo-oligomeric interfaces facilitate the hetero-octameric Mis18 complex assembly composed of 4 Mis18α, 2 Mis18ß and 2 Mis18BP1. Evaluation of structure-guided/separation-of-function mutants reveals structural determinants essential for cell cycle controlled Mis18 complex assembly and centromere maintenance. Our results provide new mechanistic insights on centromere maintenance, highlighting that while Mis18α can associate with centromeres and deposit CENP-A independently of Mis18ß, the latter is indispensable for the optimal level of CENP-A loading required for preserving the centromere identity.

Vertebrate centromere architecture: from chromatin threads to functional structures.

Centromeres are chromatin structures specialized in sister chromatid cohesion, kinetochore assembly, and microtubule attachment during chromosome segregation. The regional centromere of vertebrates consists of long regions of highly repetitive sequences occupied by the Histone H3 variant CENP-A, and which are flanked by pericentromeres. The three-dimensional organization of centromeric chromatin is paramount for its functionality and its ability to withstand spindle forces. Alongside CENP-A, key contributors to the folding of this structure include components of the Constitutive Centromere-Associated Network (CCAN), the protein CENP-B, and condensin and cohesin complexes. Despite its importance, the intricate architecture of the regional centromere of vertebrates remains largely unknown. Recent advancements in long-read sequencing, super-resolution and cryo-electron microscopy, and chromosome conformation capture techniques have significantly improved our understanding of this structure at various levels, from the linear arrangement of centromeric sequences and their epigenetic landscape to their higher-order compaction. In this review, we discuss the latest insights on centromere organization and place them in the context of recent findings describing a bipartite higher-order organization of the centromere.

Native and tagged CENP-A histones are functionally inequivalent.

BACKGROUND: Over the past several decades, the use of biochemical and fluorescent tags has elucidated mechanistic and cytological processes that would otherwise be impossible. The challenging nature of certain nuclear proteins includes low abundancy, poor antibody recognition, and transient dynamics. One approach to get around those issues is the addition of a peptide or larger protein tag to the target protein to improve enrichment, purification, and visualization. However, many of these studies were done under the assumption that tagged proteins can fully recapitulate native protein function. RESULTS: We report that when C-terminally TAP-tagged CENP-A histone variant is introduced, it undergoes altered kinetochore protein binding, differs in post-translational modifications (PTMs), utilizes histone chaperones that differ from that of native CENP-A, and can partially displace native CENP-A in human cells. Additionally, these tagged CENP-A-containing nucleosomes have reduced centromeric incorporation at early G1 phase and poorly associates with linker histone H1.5 compared to native CENP-A nucleosomes. CONCLUSIONS: These data suggest expressing tagged versions of histone variant CENP-A may result in unexpected utilization of non-native pathways, thereby altering the biological function of the histone variant.

Target Identification and Mechanistic Characterization of Indole Terpenoid Mimics: Proper Spindle Microtubule Assembly Is Essential for Cdh1-Mediated Proteolysis of CENP-A.

Centromere protein A (CENP-A), a histone H3 variant specific to centromeres, is crucial for kinetochore positioning and chromosome segregation. However, its regulatory mechanism in human cells remains incompletely understood. A structure-activity relationship (SAR) study of the cell-cycle-arresting indole terpenoid mimic JP18 leads to the discovery of two more potent analogs, (+)-6-Br-JP18 and (+)-6-Cl-JP18. Tubulin is identified as a potential cellular target of these halogenated analogs by using the drug affinity responsive target stability (DARTS) based method. X-ray crystallography analysis reveals that both molecules bind to the colchicine-binding site of ß-tubulin. Treatment of human cells with microtubule-targeting agents (MTAs), including these two compounds, results in CENP-A accumulation by destabilizing Cdh1, a co-activator of the anaphase-promoting complex/cyclosome (APC/C) E3 ubiquitin ligase. This study establishes a link between microtubule dynamics and CENP-A accumulation using small-molecule tools and highlights the role of Cdh1 in CENP-A proteolysis.

DNAJC9 prevents CENP-A mislocalization and chromosomal instability by maintaining the fidelity of histone supply chains.

The centromeric histone H3 variant CENP-A is overexpressed in many cancers. The mislocalization of CENP-A to noncentromeric regions contributes to chromosomal instability (CIN), a hallmark of cancer. However, pathways that promote or prevent CENP-A mislocalization remain poorly defined. Here, we performed a genome-wide RNAi screen for regulators of CENP-A localization which identified DNAJC9, a J-domain protein implicated in histone H3-H4 protein folding, as a factor restricting CENP-A mislocalization. Cells lacking DNAJC9 exhibit mislocalization of CENP-A throughout the genome, and CIN phenotypes. Global interactome analysis showed that DNAJC9 depletion promotes the interaction of CENP-A with the DNA-replication-associated histone chaperone MCM2. CENP-A mislocalization upon DNAJC9 depletion was dependent on MCM2, defining MCM2 as a driver of CENP-A deposition at ectopic sites when H3-H4 supply chains are disrupted. Cells depleted for histone H3.3, also exhibit CENP-A mislocalization. In summary, we have defined novel factors that prevent mislocalization of CENP-A, and demonstrated that the integrity of H3-H4 supply chains regulated by histone chaperones such as DNAJC9 restrict CENP-A mislocalization and CIN.

Artificial tethering of constitutive centromere-associated network proteins induces CENP-A deposition without Knl2 in DT40 cells.

The kinetochore is an essential structure for chromosome segregation. Although the kinetochore is usually formed on a centromere locus, it can be artificially formed at a non-centromere locus by protein tethering. An artificial kinetochore can be formed by tethering of CENP-C or CENP-I, members of the constitutive centromere-associated network (CCAN). However, how CENP-C or CENP-I recruit the centromere-specific histone CENP-A to form an artificial kinetochore remains unclear. In this study, we analyzed this issue using the tethering assay combined with an auxin-inducible degron (AID)-based knockout method in chicken DT40 cells. We found that tethering of CENP-C or CENP-I induced CENP-A incorporation at the non-centromeric locus in the absence of Knl2 (or MIS18BP1), a component of the Mis18 complex, and that Knl2 tethering recruited CENP-A in the absence of CENP-C. We also showed that CENP-C coimmunoprecipitated with HJURP, independently of Knl2. Considering these results, we propose that CENP-C recruits CENP-A by HJURP binding to form an artificial kinetochore. Our results suggest that CENP-C or CENP-I exert CENP-A recruitment activity, independently of Knl2, for artificial kinetochore formation in chicken DT40 cells. This gives us a new insight into mechanisms for CENP-A incorporation.

DNA satellite and chromatin organization at mouse centromeres and pericentromeres.

BACKGROUND: Centromeres are essential for faithful chromosome segregation during mitosis and meiosis. However, the organization of satellite DNA and chromatin at mouse centromeres and pericentromeres is poorly understood due to the challenges of assembling repetitive genomic regions. RESULTS: Using recently available PacBio long-read sequencing data from the C57BL/6 strain, we find that contrary to the previous reports of their homogeneous nature, both centromeric minor satellites and pericentromeric major satellites exhibit a high degree of variation in sequence and organization within and between arrays. While most arrays are continuous, a significant fraction is interspersed with non-satellite sequences, including transposable elements. Using chromatin immunoprecipitation sequencing (ChIP-seq), we find that the occupancy of CENP-A and H3K9me3 chromatin at centromeric and pericentric regions, respectively, is associated with increased sequence enrichment and homogeneity at these regions. The transposable elements at centromeric regions are not part of functional centromeres as they lack significant CENP-A enrichment. Furthermore, both CENP-A and H3K9me3 nucleosomes occupy minor and major satellites spanning centromeric-pericentric junctions and a low yet significant amount of CENP-A spreads locally at centromere junctions on both pericentric and telocentric sides. Finally, while H3K9me3 nucleosomes display a well-phased organization on major satellite arrays, CENP-A nucleosomes on minor satellite arrays are poorly phased. Interestingly, the homogeneous class of major satellites also phase CENP-A and H3K27me3 nucleosomes, indicating that the nucleosome phasing is an inherent property of homogeneous major satellites. CONCLUSIONS: Our findings reveal that mouse centromeres and pericentromeres display a high diversity in satellite sequence, organization, and chromatin structure.

A role for ß-1,6- and ß-1,3-glucans in kinetochore function in Saccharomyces cerevisiae.

Chromosome segregation is crucial for the faithful inheritance of DNA to the daughter cells after DNA replication. For this, the kinetochore, a megadalton protein complex, assembles on centromeric chromatin containing the histone H3 variant CENP-A, and provides a physical connection to the microtubules. Here, we report an unanticipated role for enzymes required for ß-1,6- and ß-1,3-glucan biosynthesis in regulating kinetochore function in Saccharomyces cerevisiae. These carbohydrates are the major constituents of the yeast cell wall. We found that the deletion of KRE6, which encodes a glycosylhydrolase/ transglycosidase required for ß-1,6-glucan synthesis, suppressed the centromeric defect of mutations in components of the kinetochore, foremost the NDC80 components Spc24, Spc25, the MIND component Nsl1, and Okp1, a constitutive centromere-associated network protein. Similarly, the absence of Fks1, a ß-1,3-glucan synthase, and Kre11/Trs65, a TRAPPII component, suppressed a mutation in SPC25. Genetic analysis indicates that the reduction of intracellular ß-1,6- and ß-1,3-glucans, rather than the cell wall glucan content, regulates kinetochore function. Furthermore, we found a physical interaction between Kre6 and CENP-A/Cse4 in yeast, suggesting a potential function for Kre6 in glycosylating CENP-A/Cse4 or another kinetochore protein. This work shows a moonlighting function for selected cell wall synthesis proteins in regulating kinetochore assembly, which may provide a mechanism to connect the nutritional status of the cell to cell-cycle progression and chromosome segregation.

Dynamic interplay between human alpha-satellite DNA structure and centromere functions.

Maintenance of genome stability relies on functional centromeres for correct chromosome segregation and faithful inheritance of the genetic information. The human centromere is the primary constriction within mitotic chromosomes made up of repetitive alpha-satellite DNA hierarchically organized in megabase-long arrays of near-identical higher order repeats (HORs). Centromeres are epigenetically specified by the presence of the centromere-specific histone H3 variant, CENP-A, which enables the assembly of the kinetochore for microtubule attachment. Notably, centromeric DNA is faithfully inherited as intact haplotypes from the parents to the offspring without intervening recombination, yet, outside of meiosis, centromeres are akin to common fragile sites (CFSs), manifesting crossing-overs and ongoing sequence instability. Consequences of DNA changes within the centromere are just starting to emerge, with unclear effects on intra- and inter-generational inheritance driven by centromere's essential role in kinetochore assembly. Here, we review evidence of meiotic selection operating to mitigate centromere drive, as well as recent reports on centromere damage, recombination and repair during the mitotic cell division. We propose an antagonistic pleiotropy interpretation to reconcile centromere DNA instability as both driver of aneuploidy that underlies degenerative diseases, while also potentially necessary for the maintenance of homogenized HORs for centromere function. We attempt to provide a framework for this conceptual leap taking into consideration the structural interface of centromere-kinetochore interaction and present case scenarios for its malfunctioning. Finally, we offer an integrated working model to connect DNA instability, chromatin, and structural changes with functional consequences on chromosome integrity.

Quantitative Super-Resolution Microscopy Reveals the Relationship between CENP-A Stoichiometry and Centromere Physical Size.

Centromeric chromatin is thought to play a critical role in ensuring the faithful segregation of chromosomes during mitosis. However, our understanding of this role is presently limited by our poor understanding of the structure and composition of this unique chromatin. The nucleosomal variant, CENP-A, localizes to narrow regions within the centromere, where it plays a major role in centromeric function, effectively serving as a platform on which the kinetochore is assembled. Previous work found that, within a given cell, the number of microtubules within kinetochores is essentially unchanged between CENP-A-localized regions of different physical sizes. However, it is unknown if the amount of CENP-A is also unchanged between these regions of different sizes, which would reflect a strict structural correspondence between these two key characteristics of the centromere/kinetochore assembly. Here, we used super-resolution optical microscopy to image and quantify the amount of CENP-A and DNA within human centromere chromatin. We found that the amount of CENP-A within CENP-A domains of different physical sizes is indeed the same. Further, our measurements suggest that the ratio of CENP-A- to H3-containing nucleosomes within these domains is between 8:1 and 11:1. Thus, our results not only identify an unexpectedly strict relationship between CENP-A and microtubules stoichiometries but also that the CENP-A centromeric domain is almost exclusively composed of CENP-A nucleosomes.

Humanization reveals pervasive incompatibility of yeast and human kinetochore components.

Kinetochores assemble on centromeres to drive chromosome segregation in eukaryotic cells. Humans and budding yeast share most of the structural subunits of the kinetochore, whereas protein sequences have diverged considerably. The conserved centromeric histone H3 variant, CenH3 (CENP-A in humans and Cse4 in budding yeast), marks the site for kinetochore assembly in most species. A previous effort to complement Cse4 in yeast with human CENP-A was unsuccessful; however, co-complementation with the human core nucleosome was not attempted. Previously, our lab successfully humanized the core nucleosome in yeast; however, this severely affected cellular growth. We hypothesized that yeast Cse4 is incompatible with humanized nucleosomes and that the kinetochore represented a limiting factor for efficient histone humanization. Thus, we argued that including the human CENP-A or a Cse4-CENP-A chimera might improve histone humanization and facilitate kinetochore function in humanized yeast. The opposite was true: CENP-A expression reduced histone humanization efficiency, was toxic to yeast, and disrupted cell cycle progression and kinetochore function in wild-type (WT) cells. Suppressors of CENP-A toxicity included gene deletions of subunits of 3 conserved chromatin remodeling complexes, highlighting their role in CenH3 chromatin positioning. Finally, we attempted to complement the subunits of the NDC80 kinetochore complex, individually and in combination, without success, in contrast to a previous study indicating complementation by the human NDC80/HEC1 gene. Our results suggest that limited protein sequence similarity between yeast and human components in this very complex structure leads to failure of complementation.

Sgo1 interacts with CENP-A to guide accurate chromosome segregation in mitosis.

Shugoshin-1 (Sgo1) is necessary for maintaining sister centromere cohesion and ensuring accurate chromosome segregation during mitosis. It has been reported that the localization of Sgo1 at the centromere is dependent on Bub1-mediated phosphorylation of histone H2A at T120. However, it remains uncertain whether other centromeric proteins play a role in regulating the localization and function of Sgo1 during mitosis. Here, we show that CENP-A interacts with Sgo1 and determines the localization of Sgo1 to the centromere during mitosis. Further biochemical characterization revealed that lysine and arginine residues in the C-terminal domain of Sgo1 are critical for binding CENP-A. Interestingly, the replacement of these basic amino acids with acidic amino acids perturbed the localization of Sgo1 and Aurora B to the centromere, resulting in aberrant chromosome segregation and premature chromatid separation. Taken together, these findings reveal a previously unrecognized but direct link between Sgo1 and CENP-A in centromere plasticity control and illustrate how the Sgo1-CENP-A interaction guides accurate cell division.

Nuclear lamin A-associated proteins are required for centromere assembly.

Many Lamin A-associated proteins (LAAP's) that are key constituents of the nuclear envelope (NE), assemble at the "core" domains of chromosomes during NE reformation and mitotic exit. However, the identity and function of the chromosomal core domains remain ill-defined. Here, we show that a distinct section of the core domain overlaps with the centromeres/kinetochores of chromosomes during mitotic telophase. The core domain can thus be demarcated into a kinetochore proximal core (KPC) on one side of the segregated chromosomes and the kinetochore distal core (KDC) on the opposite side, close to the central spindle. We next tested if centromere assembly is connected to NE re-formation. We find that centromere assembly is markedly perturbed after inhibiting the function of LMNA and the core-localized LAAPs, BANF1 and Emerin. We also find that the LAAPs exhibit multiple biochemical interactions with the centromere and inner kinetochore proteins. Consistent with this, normal mitotic progression and chromosome segregation was severely impeded after inhibiting LAAP function. Intriguingly, the inhibition of centromere function also interferes with the assembly of LAAP components at the core domain, suggesting a mutual dependence of LAAP and centromeres for their assembly at the core domains. Finally, we find that the localization of key proteins involved in the centromeric loading of CENP-A, including the Mis18 complex and HJURP were markedly affected in LAAP-inhibited cells. Our evidence points to a model where LAAP assembly at the core domain serves a key function in loading new copies of centromeric proteins during or immediately after mitotic exit.

Direct recruitment of Mis18 to interphase spindle pole bodies promotes CENP-A chromatin assembly.

CENP-A chromatin specifies mammalian centromere identity, and its chaperone HJURP replenishes CENP-A when recruited by the Mis18 complex (Mis18C) via M18BP1/KNL2 to CENP-C at kinetochores during interphase. However, the Mis18C recruitment mechanism remains unresolved in species lacking M18BP1, such as fission yeast. Fission yeast centromeres cluster at G2 spindle pole bodies (SPBs) when CENP-ACnp1 is replenished and where Mis18C also localizes. We show that SPBs play an unexpected role in concentrating Mis18C near centromeres through the recruitment of Mis18 by direct binding to the major SPB linker of nucleoskeleton and cytoskeleton (LINC) component Sad1. Mis18C recruitment by Sad1 is important for CENP-ACnp1 chromatin establishment and acts in parallel with a CENP-C-mediated Mis18C recruitment pathway to maintain centromeric CENP-ACnp1 but operates independently of Sad1-mediated centromere clustering. SPBs therefore provide a non-chromosomal scaffold for both Mis18C recruitment and centromere clustering during G2. This centromere-independent Mis18-SPB recruitment provides a mechanism that governs de novo CENP-ACnp1 chromatin assembly by the proximity of appropriate sequences to SPBs and highlights how nuclear spatial organization influences centromere identity.

Variation in the CENP-A sequence association landscape across diverse inbred mouse strains.

Centromeres are crucial for chromosome segregation, but their underlying sequences evolve rapidly, imposing strong selection for compensatory changes in centromere-associated kinetochore proteins to assure the stability of genome transmission. While this co-evolution is well documented between species, it remains unknown whether population-level centromere diversity leads to functional differences in kinetochore protein association. Mice (Mus musculus) exhibit remarkable variation in centromere size and sequence, but the amino acid sequence of the kinetochore protein CENP-A is conserved. Here, we apply k-mer-based analyses to CENP-A chromatin profiling data from diverse inbred mouse strains to investigate the interplay between centromere variation and kinetochore protein sequence association. We show that centromere sequence diversity is associated with strain-level differences in both CENP-A positioning and sequence preference along the mouse core centromere satellite. Our findings reveal intraspecies sequence-dependent differences in CENP-A/centromere association and open additional perspectives for understanding centromere-mediated variation in genome stability.

The CENP-A nucleosome: where and when it happens during the inner kinetochore's assembly.

CENP-A is an essential histone variant that replaces the canonical H3 at the centromeres and marks these regions epigenetically. The CENP-A nucleosome is the specific building block of centromeric chromatin, and it is recognized by CENP-C and CENP-N, two components of the constitutive centromere-associated network (CCAN), the first protein layer of the kinetochore. Recent proposals of the yeast and human (h)CCAN structures position the assembly on exposed DNA, suggesting an elusive spatiotemporal recognition. We summarize the data on the structural organization of the CENP-A nucleosome and the binding of CENP-C and CENP-N. The latter posits an apparent contradiction in engaging the CENP-A nucleosome versus the CCAN. We propose a reconciliatory model for the assembly of CCAN on centromeric chromatin.

DNA satellite and chromatin organization at house mouse centromeres and pericentromeres.

Centromeres are essential for faithful chromosome segregation during mitosis and meiosis. However, the organization of satellite DNA and chromatin at mouse centromeres and pericentromeres is poorly understood due to the challenges of sequencing and assembling repetitive genomic regions. Using recently available PacBio long-read sequencing data from the C57BL/6 strain and chromatin profiling, we found that contrary to the previous reports of their highly homogeneous nature, centromeric and pericentromeric satellites display varied sequences and organization. We find that both centromeric minor satellites and pericentromeric major satellites exhibited sequence variations within and between arrays. While most arrays are continuous, a significant fraction is interspersed with non-satellite sequences, including transposable elements. Additionally, we investigated CENP-A and H3K9me3 chromatin organization at centromeres and pericentromeres using Chromatin immunoprecipitation sequencing (ChIP-seq). We found that the occupancy of CENP-A and H3K9me3 chromatin at centromeric and pericentric regions, respectively, is associated with increased sequence abundance and homogeneity at these regions. Furthermore, the transposable elements at centromeric regions are not part of functional centromeres as they lack CENP-A enrichment. Finally, we found that while H3K9me3 nucleosomes display a well-phased organization on major satellite arrays, CENP-A nucleosomes on minor satellite arrays lack phased organization. Interestingly, the homogeneous class of major satellites phase CENP-A and H3K27me3 nucleosomes as well, indicating that the nucleosome phasing is an inherent property of homogeneous major satellites. Overall, our findings reveal that house mouse centromeres and pericentromeres, which were previously thought to be highly homogenous, display significant diversity in satellite sequence, organization, and chromatin structure.

Conserved and divergent mechanisms of inner kinetochore assembly onto centromeric chromatin.

Kinetochores are large protein complexes built on centromeric chromatin that mediate chromosome segregation. The inner kinetochore, or constitutive centromere-associated network (CCAN), assembles onto centromeres defined by centromere protein A (CENP-A) nucleosomes (CENP-ANuc), and acts as a platform for the regulated assembly of the microtubule-binding outer kinetochore. Recent cryo-EM work revealed structural conservation of CCAN, from the repeating human regional centromeres to the point centromere of budding yeast. Centromere recognition is determined mainly through engagement of duplex DNA proximal to the CENP-A nucleosome by a DNA-binding CENP-LN channel located at the core of CCAN. Additional DNA interactions formed by other CCAN modules create an enclosed DNA-binding chamber. This configuration explains how kinetochores maintain their tight grip on centromeric DNA to withstand the forces of chromosome segregation. Defining the higher-order architecture of complete kinetochore assemblies with implications for understanding the 3D organisation of regional centromeres and mechanisms of kinetochore dynamics, including how kinetochores sense and respond to tension, are important future directions.

SUMO control of centromere homeostasis.

Centromeres are unique chromosomal loci that form the anchorage point for the mitotic spindle during mitosis and meiosis. Their position and function are specified by a unique chromatin domain featuring the histone H3 variant CENP-A. While typically formed on centromeric satellite arrays, CENP-A nucleosomes are maintained and assembled by a strong self-templated feedback mechanism that can propagate centromeres even at non-canonical sites. Central to the epigenetic chromatin-based transmission of centromeres is the stable inheritance of CENP-A nucleosomes. While long-lived at centromeres, CENP-A can turn over rapidly at non-centromeric sites and even erode from centromeres in non-dividing cells. Recently, SUMO modification of the centromere complex has come to the forefront as a mediator of centromere complex stability, including CENP-A chromatin. We review evidence from different models and discuss the emerging view that limited SUMOylation appears to play a constructive role in centromere complex formation, while polySUMOylation drives complex turnover. The deSUMOylase SENP6/Ulp2 and the proteins segregase p97/Cdc48 constitute the dominant opposing forces that balance CENP-A chromatin stability. This balance may be key to ensuring proper kinetochore strength at the centromere while preventing ectopic centromere formation.