Skip to main page content
U.S. flag
See More

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2025 Oct 14;53(19):gkaf1038.
doi: 10.1093/nar/gkaf1038.

Direct observation of interdependent and hierarchical kinetochore assembly on individual centromeres

Changkun Hu  1 , Andrew R Popchock  1 , Angelica Andrade Latino  1 , Charles L Asbury  2 , Sue Biggins  1
Affiliations

Direct observation of interdependent and hierarchical kinetochore assembly on individual centromeres

Changkun Hu et al. Nucleic Acids Res. .

Abstract

Kinetochores are megadalton protein machines that harness microtubules to segregate chromosomes during cell division. The kinetochores must assemble after DNA replication during every cell cycle onto specialized regions of chromosomes called centromeres, but the order and regulation of their assembly remains unclear due to the complexity of kinetochore composition and the difficulty resolving individual kinetochores in vivo. Here, by adapting a prior single-molecule method for monitoring kinetochore assembly in budding yeast lysates, we identify a sequential order of assembly and uncover previously unknown interdependencies between subcomplexes. We show that inner kinetochore assembly depends partly on outer kinetochore components, and that outer kinetochore branches do not assemble independently of one another. Notably, Mif2 assembly is a rate-limiting step that can be accelerated by binding to the Mtw1 subcomplex, thereby promoting rapid assembly of many inner and outer kinetochore components. The importance of controlling kinetochore assembly kinetics is supported by a Mif2 mutant lacking both autoinhibition and Mtw1 subcomplex binding activity, which leads to defective kinetochore-microtubule attachments when the centromeric histone variant Cse4 is overexpressed. Altogether, our work provides a direct view of kinetochore assembly and reveals highly interdependent regulatory events that control its order and timing.

Plain language summary

Kinetochores are large protein machines containing >50 proteins that are vital for moving chromosomes to daughter cells when they divide. Kinetochores must assemble every cell cycle after DNA replication, but the order and regulation of their assembly remains unclear. Here, a single-molecule microscopy method was adapted to watch kinetochores assemble over time. The new approach enabled the sequential order of assembly to be mapped and revealed previously unknown interdependencies between kinetochore components. It also identified an unknown mechanism that cells use to relieve autoinhibition of kinetochore assembly. Together, the work provides an unprecedented view of the kinetochore assembly process and the regulatory events that control its order and timing.

PubMed Disclaimer

Conflict of interest statement

None declared.

Figures

Graphical Abstract
Graphical Abstract
Figure 1.
Figure 1.
Total internal reflection fluorescence microscopy (TIRFM) assay reveals the assembly timing of kinetochore subcomplexes. (A) Schematic of the budding yeast kinetochore based on [16] demonstrating known interactions between inner kinetochore and outer kinetochore. Only proteins included in this study are labeled. Note that human homolog names are CENP-A, CENP-C, CENP-HIK, CENP-LN CENP-TW, CENP-OP, and CENP-QU (See Table 1). (B) Schematic describing time series TIRFM assay. Cells that contained fluorescent, GFP-tagged kinetochore proteins were treated with benomyl to depolymerize microtubules and arrest cells in mitosis. Lysates were prepared and then introduced into a flow-channel with fluorescent, Atto 647-tagged CEN3 DNA tethered sparsely onto the coverslip surface. After incubation for various times (0, 5, 10, 30, 60, 90, 120, 150, or 180 min), the lysates were washed out and then TIRFM was performed. Created in BioRender [Hu, C. (2025; https://BioRender.com/b83s784) and Hu, C. (2024; https://BioRender.com/n88v756)]. (C) Representative TIRFM images of surface-tethered wild-type or mutant CEN3 DNA molecules (at locations indicated by yellow circles) after incubation for the indicated times with lysates from strains carrying Ndc10-GFP (SBY22903) or Mif2-GFP (SBY22094). Ndc10-GFP accumulated rapidly and specifically on wild-type CEN3 DNAs, whereas Mif2-GFP accumulated more slowly (GFP foci shown in grayscale). (D) Percentages of CEN3 or CEN3 mutant DNAs with colocalized Ndc10-GFP (SBY22903) or Mif2-GFP (SBY22094) versus time. Error bars represent the standard deviation over three biological repeats. At least 3000 DNA molecules were imaged for each time point from each biological replicate. Fitting curves were generated using two-parameter kinetic equations as detailed in the ‘Materials and methods’ section.
Figure 2.
Figure 2.
Centromeric nucleosome assembly is followed by incorporation of the CCAN subcomplexes. (A) Percentages of CEN3 DNAs with colocalized GFP-tagged Ndc10 (SBY22903), Cse4 (SBY22195), Ame1 (SBY22119), Mif2 (SBY22094), Ctf19 (SBY22116), Wip1 (SBY22207), Iml3 (SBY22199), or Ctf3 (SBY22203). Ndc10-GFP and Mif2-GFP data are replotted from Fig. 1C for comparison. Error bars represent the standard deviation over three biological repeats. At least 3000 DNA molecules were imaged for each time point from each biological replicate. Control experiments using the assembly-blocking CEN3 mutant DNA confirmed that all of these GFP-tagged proteins assembled specifically on wild-type CEN3 DNA (Supplemental Fig. S2). T30 is labeled for each GFP-tagged protein. (B) Schematic model illustrating the assembly order of the inner kinetochore. (i) Ndc10 binds to CEN3 DNA to initiate kinetochore assembly. (ii) The Cse4 centromeric nucleosome forms. (iii) Ame1 binds to the centromeric nucleosome. (iv) Mif2 and the nonessential CCAN proteins assemble. T30 is labeled for each assembly step.
Figure 3.
Figure 3.
Relieving autoinhibition of the Mtw1 subcomplex with dsn1-2D improves outer kinetochore assembly. (A) Left: Percentages of CEN3 DNAs with colocalized Ndc10-GFP (SBY22903) and GFP-tagged outer kinetochore proteins Dsn1 (SBY22153), Nuf2 (SBY23256), Kre28 (SBY24188), and Stu2 (SBY22135). Ndc10-GFP data are replotted from Figs 1C and 2A for comparison. Right: Same data with zoom-in left axis. (B) Schematic illustrating the regulation of the Mtw1 subcomplex by Aurora B phosphorylation. Mtw1 binding to Mif2 is autoinhibited (“closed”, top panel) until Dsn1 is phosphorylated by the Aurora B kinase (“open”, bottom panel). Once open, Mtw1 can bind to the N-terminus of Mif2. (C) Percentages of CEN3 DNAs with colocalized outer kinetochore proteins Dsn1-2D-GFP (SBY22159), Nuf2-GFP (SBY23258), and Kre28-GFP (SBY24190) in the dsn1-2D background. (D) Percentages of maximum CEN3 DNAs with colocalized GFP-tagged kinetochore proteins after assembly in the indicated lysates, including the same data from Fig. 3A and C plus data for cnn1Δ DSN1-GFP (SBY24769) and cnn1Δ dsn1-2D-GFP (SBY24747). Maximum colocalization was obtained from the Cmax or 300 min after incubation when the Cmax is unavailable. All error bars represent the standard deviation over three biological repeats. At least 3000 DNA molecules were imaged for each time point from each biological replicate.
Figure 4.
Figure 4.
Relieving autoinhibition of the Mtw1 subcomplex with dsn1-2D improves assembly of CCAN components. (A  –F) Percentages of CEN3 DNAs with colocalized GFP-tagged inner kinetochore proteins after assembly in DSN1 and dsn1-2D lysates. All DSN1 data are replotted from Fig. 2A for comparison. GFP-tagged proteins in the dsn1-2D background include Ndc10 (SBY22905), Cse4 (SBY22193), Ame1 (SBY22117), Ctf19 (SBY22114), Iml3 (SBY22197), and Ctf3 (SBY22201). (G) Percentages of CEN3 DNAs with colocalized GFP tagged inner kinetochore proteins after assembly in dsn1-2D lysates, including the same data from panels (A)–(F), plus data for Mif2-GFP (SBY22092) and Wip1 (SBY22205). T30 is labeled for each GFP-tagged protein. (H) Schematic model illustrating the change in assembly order of the inner kinetochore in dsn1-2D lysates (right pathway) compared to wild type (left pathway, from Fig. 2B). (i) Ndc10 binds to CEN3 DNA to initiate kinetochore assembly. (ii) The Cse4 centromeric nucleosome forms. (iii) Mif2 binds to the centromeric nucleosome. (iv) Ame1 and the rest of nonessential CCAN proteins (not included) assemble.T30 is labeled for each assembly step. All error bars represent the standard deviation over three biological repeats. At least 3000 DNA molecules were imaged for each time point from each biological replicate.
Figure 5.
Figure 5.
Relieving autoinhibition of the Mtw1 subcomplex with dsn1-2D also alleviates Mif2 autoinhibition. (A) Percentages of CEN3 DNAs with colocalized Mif2-GFP after assembly in DSN1 (SBY22094, data replotted from Fig. 2A) or dsn1-2D lysates (SBY22092, data replotted from Fig. 4G). (B) Percentages of CEN3 DNAs with colocalized Mif2-GFP after assembly in lysates made from a dsn1-AID strain (SBY22872) treated with auxin to deplete the Dsn1-AID protein, or with DMSO as a control. (C) Schematics illustrating Mif2 autoinhibition. (i) Mif2 autoinhibition (closed) is relieved by binding to Cse4 (left pathway), making Mif2-N available (open) for binding to Mtw1c. Mif2 autoinhibition can also be relieved by binding to Mtw1c (right pathway), making Mif2 available for Cse4 binding (open). (ii, iii) Deleting the N-terminus of Mif2 (Mif2-ΔN, ii) or replacing it with an N-terminal fragment of Ame1 (Mif2Ame1N, iii) can relieve Mif2 autoinhibition. When the Ame1 N-terminus is grafted onto Mif2, it also restores binding to Mtw1c. (D) Percentages of CEN3 DNAs with colocalized wild-type Mif2-GFP or colocalized truncation mutant Mif2-ΔN-GFP after assembly in either DSN1 or dsn1-2D lysates (from strains SBY22094, SBY23248, SBY22092, and SBY23250). Data from strains with wild-type Mif2-GFP are replotted from Figs 1C and 4G. (E) Percentages of CEN3 DNA with colocalized wild-type Mif2-GFP or colocalized swap mutant Mif2Ame1N-GFP after assembly in either DSN1 or dsn1-2D lysates (from strains SBY22094, SBY23252, SBY22092, and SBY23254). Data from strains with wild-type Mif2-GFP are replotted from Figs 1C and 4G. All error bars represent the standard deviation over three biological repeats. At least 3000 DNA molecules were imaged for each time point from each biological replicate.
Figure 6.
Figure 6.
Interactions between Mif2 and the Mtw1 subcomplex improve assembly of nonessential CCAN components. (A) Percentages of CEN3 DNAs with colocalized Ctf19-GFP in wild type (SBY22116), mif2-ΔN (SBY24438), and mif2AME1N(SBY24420) lysates. (B) Percentages of CEN3 DNAs with colocalized Iml3-GFP in wild type (SBY22199), mif2-ΔN (SBY24436), and mif2AME1N (SBY24418) lysates. (C) Percentages of CEN3 DNAs with colocalized Ctf3-GFP in wild type (SBY22203), mif2-ΔN (SBY24426), and mif2AME1N (SBY24422) lysates. (D) Percentage of CEN3 DNA with colocalized Wip1-GFP in wild type (SBY22207), mif2-ΔN (SBY24430), and mif2AME1N (SBY24424) lysates. All data from strains with wild-type Mif2 are replotted from Fig. 2A for comparison. Error bars represent the standard deviation over three biological repeats. At least 3000 DNA molecules were imaged for each time point from each biological replicate.
Figure 7.
Figure 7.
Cse4 overexpression in mif2-ΔN cells generates defective kinetochore-microtubule attachments. (A) Five-fold serial dilutions of MIF2-GFP (SBY22095), mif2-ΔN (SBY23249), pGAL-CSE4 MIF2-GFP (SBY24220), and pGAL-CSE4 mif2-ΔN-GFP (SBY24222) yeast strains. Cells were grown on glucose and galactose plates and incubated at 23°C for 48 h. (B) Representative images of Mif2-GFP, Mif2-ΔN-GFP, and Ndc10-mCherry in wild type (SBY24202), mif2-ΔN (SBY24204), pGAL-CSE4 MIF2-GFP (SBY24212), and pGAL-CSE4 mif2-ΔN-GFP (SBY24216) yeast strains. Cells were arrested in G1 with alpha factor and released into galactose media to induce Cse4 overexpression and harvested after 120 min. The borders of the cells are outlined in white. The mCherry signal is weak in the merged images so the merges appear green. (C) Percentages of cells with more than two GFP foci in strains from panel (B) at different time points. At least 100 cells were imaged for each time point over three biological repeats. Error bars are standard deviation over three biological repeats. (D) Percentages of cells in anaphase determined by DAPI staining in strains from panel (A) and pGAL-CSE4 mif2-ΔN mad3Δ (SBY24224). At least 100 cells were imaged over three biological repeats for each time point. Error bars are standard deviation over three biological repeats. All error bars represent the standard deviation over three biological repeats.
Figure 8.
Figure 8.
Summary of assembly order and interdependence between outer and inner kinetochore subcomplexes. (A) Schematic showing the order of assembly of indicated kinetochore components. Top row: Ndc10 and the centromeric histone H3 variant Cse4 assemble first, followed by the essential CCAN components, Ame1 and Mif2, and then the nonessential CCAN components. Bottom row: Among outer kinetochore components, Mtw1c assembles first, followed by Ndc80c, and then the KNL subcomplex and Stu2. (B) Unexpected kinetochore assembly dependencies discovered using a single molecule de novo assembly assay. Dsn1-2D increased the assembly efficiency of nonessential CCAN components and accelerated Mif2 recruitment, in addition to its known role in promoting outer kinetochore assembly. Cnn1 enhanced the assembly of Dsn1-2D into the kinetochore.
See this image and copyright information in PMC

Update of

References

    1. Sundararajan K, Straight AF Centromere identity and the regulation of chromosome segregation. Front Cell Dev Biol. 2022; 10:914249. 10.3389/fcell.2022.914249. - DOI - PMC - PubMed
    1. Thomas GE, Renjith MR, Manna TK Kinetochore-microtubule interactions in chromosome segregation: lessons from yeast and mammalian cells. Biochem J. 2017; 474:3559–77. 10.1042/BCJ20170518. - DOI - PubMed
    1. Shukla S, Kumar A Centromere chromatin dynamics at a glance. Epigenomes. 2022; 6:39–49. 10.3390/epigenomes6040039. - DOI - PMC - PubMed
    1. Stirpe A, Heun P The ins and outs of CENP-A: chromatin dynamics of the centromere-specific histone. Semin Cell Dev Biol. 2023; 135:24–34. 10.1016/j.semcdb.2022.04.003. - DOI - PubMed
    1. McAinsh AD, Marston AL The four causes: the functional architecture of centromeres and kinetochores. Annu Rev Genet. 2022; 56:279–314. 10.1146/annurev-genet-072820-034559. - DOI - PMC - PubMed

MeSH terms

Substances

LinkOut - more resources