Ensemble-Level Organization of Human Kinetochores and Evidence for Distinct Tension and Attachment Sensors

Abstract

Highlights d 3D mapping of kinetochore architecture in human RPE1 cells d Outer kinetochore (NDC80, Mad1, and RZZ) has high nematic order d NDC80 jackknives and KNL1 unravels upon loss of attachment and tension, respectively d Recruitment of Mad2 is only coupled to the occupancy sensor (NDC80)

Cyclin B1 scaffolds MAD 1 at the kinetochore corona to activate the mitotic checkpoint

“…In this arrangement, MAD1 would form the bridge in a tripartite Cyclin B1:MAD1:BUB1 complex that connects the corona with the KMN network. We currently favour this latter model because very recent spatial positioning data on MAD1‐pT716 and MAD2 demonstrates that these signals move slightly outwards towards the RZZ complex after BUB1 depletion (preprint: Roscioli et al , ). It will be important in future to determine whether MAD1 bridges the corona and KMN network, because if it does, then this would have important implications for both SAC signalling and corona formation.…”

Aurora B Tension Sensing Mechanisms in the Kinetochore Ensure Accurate Chromosome Segregation

“…Distance measurements from inner centromere to outer kinetochore suggest these regions may be separated by up to 100 nm [58,126], making the proposed 80 nm dog leash [94] insufficient for Aurora B to reach the outer kinetochore even in tensionless scenarios. If tethered in the inner kinetochore near CENP-C [102] or CENP-A [146], a 80 nm leash would allow Aurora B to reach the outer kinetochore in the absence of tension, but be excluded under tension based on studies that show significant intra-kinetochore stretch:~20 nm deformation in CENP-C regions [60,75],~40 nm jack-knifing action in Ndc80, and~50 nm unraveling of Knl1 [75]. The kinetochore localization model could also be combined with tension-sensitive models; Aurora B activity could be regulated by tension sensors located in the kinetochore.…”

“…A population of INCENP independent of other CPC members interacts directly with the COMA inner kinetochore sub complex and recruits Aurora B [123,124], and thus could regulate kinase activity at the kinetochore location. In addition to INCENP, Knl1 is a proposed tension-sensitive Aurora B regulator [75]. Phospho-regulation of Aurora B activity in the kinetochore-localized pool has also been demonstrated; only the kinetochore-based pool is phosphorylated by the Chk2 kinase and this phosphorylation confers kinase activity to respond to unattached kinetochores and activate the spindle checkpoint [146,147].…”

“…In human RPE1 cells, stretching between inner (CENP-A) and outer (Ndc80) components increases throughout prometaphase with 40% of kinetochores visibly stretched in metaphase [74]. The outer kinetochore experiences reorganization under tension via a jackknifing action of the Ndc80 protein, unraveling of the Knl1 protein [75], and extension of the unstructured domain of CENP-T [74]. Recent microscopy and modeling of CenpA-Ndc80 distances in different species suggests that at least 50% of intra-kinetochore distance changes is caused by malleable reorganization of proteins as well as elastic stretching of proteins [76].…”

“…There is support for Aurora B recruitment by kinetochore proteins; in yeast, the COMA kinetochore sub-complex has been shown to physically interact with INCENP and recruit Aurora B to inner kinetochores independently of Survivin and other CPC members on chromatin [123]. Additionally, the kinetochore protein Knl1 is required for Aurora B activity [148] and Knl1 has been shown recently to change structural conformation under tension [75], acting perhaps as a binding regulator of Aurora B [38,102].…”