Kinetochore clustering depends on complex protein interactions

Proper cell division depends on faithful segregation of genetic information into daughter cells. After DNA is duplicated in a dividing cell, it is condensed and packaged into chromosomes. Post-replication chromosomes consist of two sister chromatids connected by cohesin. Sister chromatids are eventually pulled apart by microtubules, which grab onto specialized protein structures that form at the centromere called kinetochores.  In mammals, a structure called the fibrous corona forms around the outer kinetochore to facilitate microtubule capture. Formation of the fibrous corona is triggered by phosphorylation by the Mps1 protein. In simpler eukaryotes like budding yeast, this process looks a bit different. “Yeasts do not form fibrous coronas on their unattached kinetochores. However, like animal cells, they recruit additional microtubule binding proteins to unattached kinetochores and these proteins cluster the kinetochores to promote microtubule binding,” explains Dr. Darren Mallett. The Stu1 and Slk19 proteins are required for this process. Stu1 and Slk19 are found at unattached kinetochores, and Mps1 activity is required for Stu1 to localize to the kinetochores. Despite the well-characterized roles for Stu1 and Slk19 in the yeast cell cycle, how exactly they promote kinetochore clustering was unclear. Researchers in the Biggins lab in the Basic Sciences Division sought to characterize this mechanism and precisely define the role of Mps1 in this process.

The team, led by Mallett, wanted to get a clearer picture of how unattached kinetochore clustering works. First, they sought to define the region of Stu1 that interacts with kinetochores. They expressed FLAG-tagged Stu1 proteins lacking either the TOG1 or CL domains and isolated the proteins from yeast. They found that Stu1 proteins lacking the TOG1 domain had no interactions with a core kinetochore protein or a key spindle checkpoint protein, indicating that the TOG1 domain interacts directly with the kinetochore. Deletion of the CL domain had no impact on interactions with the kinetochore and spindle checkpoint proteins. Deletion of either domain severely reduced the Stu1-Slk19 protein interaction, a surprising result given that previous work reported that only the CL domain was required for this interaction. They next created a mutation in the TOG1 domain to abolish kinetochore binding while leaving the TOG1 domain largely intact. They used this mutant, called Stu1^∆K94, to further interrogate the role of Stu1 in kinetochore clustering.

To understand whether Stu1 interactions with kinetochore proteins are required for clustering, the group synchronized wild type yeast, yeast with the Stu1^∆K94 mutation, and yeast lacking Slk19 in the G1 phase of the cell cycle and released them into media containing a drug that generates unattached kinetochores. All three yeast strains expressed fluorescently marked kinetochores. In wild type yeast, they only saw two or three kinetochore foci, indicating that there was robust kinetochore clustering. In yeast lacking Slk19, they saw more kinetochore foci, indicating that the mutation was destabilizing the kinetochore clusters. They saw similar results in the yeast expressing Stu1^∆K94. Next, they tested whether the Stu1^∆K94 mutant was required to recruit Slk19 to the kinetochores. They found that there was a 76% reduction in Slk19 at the unattached kinetochore in Stu1^∆K94 mutants, indicating that the Stu1 TOG1 kinetochore binding capacity is required for Slk19 recruitment and robust kinetochore clustering.

Even though it is established that Mps1 activity is necessary for Stu1 localization to the kinetochore, it was unclear whether Mps1 was required for unattached kinetochore clustering. To test this, they counted the number of fluorescent kinetochores in wild type cells and cells lacking Mps1 protein. They found that the cells lacking Mps1 had significantly more kinetochore foci, indicating that Mps1 activity is required for kinetochore clustering.

Mps1 also controls the mitotic spindle checkpoint that ensures microtubules are attached correctly to kinetochores before cell division begins, and the researchers wanted to know if kinetochore clustering impacted the spindle checkpoint. To test if the spindle checkpoint was required for kinetochore clustering, they counted the number of kinetochore foci in cells lacking key proteins required for the spindle checkpoint, but they did not detect differences in the number of foci between these cells and wild type controls. To test if kinetochore clustering is required for spindle checkpoint activation, they analyzed spindle checkpoint activity in the cells expressing Stu1^∆K94 or lacking Slk19, but they did not detect any differences in checkpoint activation between their mutant cells and wild type controls. These results indicate that the spindle checkpoint and kinetochore clustering are distinct and that Mps1 has independent roles in controlling both pathways.

Next, the team sought to define whether Mps1 is necessary for the Stu1-Slk19 interaction. They purified Stu1 from wild type and cells lacking Mps1 and saw that Slk19 was only bound to Stu1 isolated from wild type cells, indicating that Mps1 is necessary for the two proteins to interact. Because Mps1 is a kinase, they thought that it may be phosphorylating a specific residue on Stu1 to promote the Stu1-Slk19 interaction. They identified two MELT motifs in Stu1 using sequence alignment strategies. “This was a key result that helped us narrow down that Mps1 was required,” explains Mallett. Mps1 is known to phosphorylate MELT motifs in other proteins, so they hypothesized Mps1-driven MELT phosphorylation may be driving the Stu1-Slk19 interaction. To test this biochemically, they expressed a Stu1 mutant protein with MELT motifs that are unable to be phosphorylated and purified it from cells to see if Slk19 could still interact. They saw no Slk19 interactions with their un-phosphorylated Stu1 mutant, indicating that MELT phosphorylation is necessary for the Stu1-Slk19 interaction. In vivo, they saw that cells expressing the un-phosphorylated Stu1 mutant had less Slk19 present at the unattached kinetochores and more kinetochore foci, indicating that MELT phosphorylation is necessary for Slk19 localization and unattached kinetochore clustering.

Previous work had shown that Stu1 overexpression leads to the formation of large structures surrounding the kinetochores, leading researchers to think that Stu1 and Slk19 may form oligomeric networks around unattached kinetochores. Mallett and Biggins hypothesized that a larger network could explain why disrupting MELT phosphorylation also disrupts Stu1 localization to the kinetochore, even though they established the TOG1 as the kinetochore-binding domain. Mengqiu Jiang used cryo-electron microscopy to look at the structure of the Stu1-Slk19 complex. They observed several string-like filaments comprised of globular proteins and coiled-coil proteins. AlphaFold3 modeling predicted that the globular proteins were Stu1 and the coiled-coil proteins were Slk19. While they were unable to solve the precise structure of the protein complex due to technical limitations, these results show that Slk19 exists as an elongated, coiled protein, and that the Stu1-Slk19 complex can form long, stringy filaments to promote unattached kinetochore clustering.

Integrating all of this data, the group proposes a new model for unattached kinetochore clustering. First, TOG1 binds kinetochores, placing Stu1 in a favorable position for the MELT motifs to be phosphorylated by Mps1. The authors think MELT phosphorylation could initiate Stu1 and Slk19 oligomerization, or Slk19 could bind Stu1 MELTs to further stabilize the TOG1-kinetochore interactions. From there, Stu1 and Slk19 form stringy filaments that promote unattached kinetochore clustering. In addition to mechanistically defining how Stu1-Slk19 promote unattached kinetochore clustering, this work also highlights a new role for Mps1 in cell cycle regulation. “One of the most striking findings of this work is that Mps1, one of the key enzymes involved in building the fibrous corona, also builds the Stu1-Slk19 complex on unattached kinetochores in yeast. So, it appears that Mps1 has a conserved function of helping with initial kinetochore attachment, but it uses different mechanisms for different lineages,” explains Mallett. In the future, the authors hope that these results can be leveraged by others in the field working to understand the molecular details of kinetochore clustering.

This work was supported by funding from the National Institutes of Health and the Howard Hughes Medical Institute.

Fred Hutch/University of Washington/Seattle Children’s Cancer Consortium Member Dr. Sue Biggins contributed to this work.

Mallett DR, Jiang M, Minnuto GM, Biggins S. 2026. Kinetochore clustering is mediated by Mps1 phosphorylation of conserved MELT motifs in Stu1. J Cell Biol. 225(9):e202601143. doi: 10.1083/jcb.202601143.