During vertebrate mitosis, the centromere-associated kinesin CENP-E (centromere protein E) transports misaligned chromosomes to the plus ends of spindle microtubules. Subsequently, the kinetochores that form at the centromeres establish stable associations with microtubule ends, which assemble and disassemble dynamically. Here we provide evidence that after chromosomes have congressed and bi-oriented, the CENP-E motor continues to play an active role at kinetochores, enhancing their links with dynamic microtubule ends. Using a combination of single-molecule approaches and laser trapping in vitro, we demonstrate that once reaching microtubule ends, CENP-E converts from a lateral transporter into a microtubule tip-tracker that maintains association with both assembling and disassembling microtubule tips. Computational modelling of this behaviour supports our proposal that CENP-E tip-tracks bi-directionally through a tethered motor mechanism, which relies on both the motor and tail domains of CENP-E. Our results provide a molecular framework for the contribution of CENP-E to the stability of attachments between kinetochores and dynamic microtubule ends.
Access optionsAccess options
Subscribe to Journal
Get full journal access for 1 year
only $18.75 per issue
All prices are NET prices.
VAT will be added later in the checkout.
Rent or Buy article
Get time limited or full article access on ReadCube.
All prices are NET prices.
We thank members of the Grishchuk and Ataullakhanov laboratories for stimulating discussions and technical assistance, A. Korbalev for help with the model videos, M. Porter for a kind gift of axonemes, M. Ostap, M. Lampson and J. R. McIntosh for critical reading of the manuscript, and E. Ballister and M. Lampson for assistance with live cell imaging. We thank University of California, San Diego, Neuroscience Microscopy Shared Facility (P30 NS047101) and J. R. McIntosh for help with the initial phase of this work (GM 033787). This work has been supported by grants from the NIH to E.L.G. (R01-GM098389) and to D.W.C. (R01-GM29513), and grants to F.I.A from Russian Fund for Basic Research (12-04-00111-a, 13-04-40188-H and RFBR 13-04-40190-H) and Presidium of Russian Academy of Sciences (Mechanisms of the Molecular Systems Integration and Molecular and Cell Biology programmes). B.V. has been supported by a postdoctoral fellowship from the Human Frontiers Science Program. E.L.G. is a Kimmel Scholar; her work is supported in part by the Pennsylvania Muscle Institute. D.W.C. receives salary support from the Ludwig Institute for Cancer Research.
Integrated supplementary information
Interactions between CENP-E and microtubule lattice, as determined with single molecule TIRF and bead assays. Here and in other tables the numbers are Mean±SEM, N, where N is the total number of observed events. For the beads assay only data from the microscopy chambers in which no more than 30% beads were motile are included. When bead motility is that infrequent, more than 80% of moving beads are likely to be driven by single motor molecules53. The velocity of single truncated CENP-E molecules reported here is consistent with findings in28,30, but it is faster than in6 due to the differences in assay conditions.
Interactions between CENP-E and microtubule tips. Data reported here were obtained with complexes containing 1-2 CENP-E dimers. The association time with stable tips was examined using Taxol-stabilized microtubules. The association time and travel distance with polymerizing tips were determined from single exponential fits of the respective histograms (Supplementary Fig. S4 panels a,c, red curves). The association time and travel distance with the depolymerizing tips (Supplementary Fig. S4 panels b,d) and the association times with stable microtubule tips are reported as mean values. Time and distance of full-length CENP-E motility with microtubule disassembly are likely to be underestimated because they were frequently limited by a short microtubule length. NA– not applicable.
Model parameters. Our mathematical model contains 11 independent parameters: 5 kinetic parameters to describe the rate of the transitions between different states, as illustrated in Fig. 7a,b, and 6 mechanical parameters. For further details about specifying the values of model parameters and model calibration see Mathematical Model of CENP-E Motility in Supplementary Note.
HeLa cells expressing Tubulin–EYFP and Histone H2B–RFP (natural colours are inverted in this video for better visualization) were released from the metaphase arrest in the presence (upper panel) or absence (lower panel) of the CENP-E inhibitor GSK-923295 at the beginning of these sequences, played at 4 fps, 480x faster than recorded. In bi-polar spindles with rigor-locked CENP-E, several chromosomes lose their alignment and become stuck at the poles, leading to a prolonged cell cycle arrest.
At the start of this sequence, a 0.5 μm bead coated with TR CENP-E is held with a weak optical trap near a coverslip-attached Taxol-stabilized microtubule. A piezo-stage is then repositioned to promote the bead’s contact with the microtubule. When the bead starts walking, the trap is turned off to record the velocity and run length of the bead motions. In such assays, the TR-coated beads moved similarly to the beads coated with FL CENP-E, demonstrating that the non-motor domains have little effect on CENP-E traditional transport along a microtubule wall. Video was recorded in DIC at 7 fps and 100 millisec exposure, and played 2x faster than recorded. Scale bar, 2 μm.
Two TR CENP-E dimers sequentially land on the microtubule lattice and quickly walk to its growing plus end, where they detach readily. The microtubule then starts shortening, while another TR motor lands, walks and falls off the disassembling tip. Images were acquired with two-colour TIRF, so the microtubule is dimmer towards the plus end where it is farther away from the coverslip. The coverslip-attached Rhodamine-labeled microtubule seed is green, because Rhodamine is excited with a 488 nm laser. The 640 nm laser was used to visualize ‘red’ Hilyte-647 tubulin. Time-lapse image sequence was acquired at 2 fps, played 5x faster than recorded. Scale bar, 2 μm.
A single FL CENP-E dimer reaches and then tracks with a disassembling microtubule end, exhibiting processive motility in both plus- and minus-end directions. Motions were recorded as in Supplementary Video S3 but continuously at 100 millisec per channel, and played 5x faster than recorded. Scale bar, 2 μm.
A FL CENP-E tetramer lands on the microtubule lattice, reaches a growing tip and stays tip-associated during the phases of growth and shortening. Another FL CENP-E complex, likely to be a single dimer, joins the first complex as it tracks with the disassembling end, until both get stuck to the coverslip surface (see Supplementary Fig. S3b, left panel, for the respective kymograph). Motions were recorded as in Supplementary Video S4, and played 3.3x faster than recorded. Scale bar, 2 μm.
Arrow points to a single FL CENP-E dimer walking to a growing microtubule tip. Subsequently, the arrow’s position is fixed; the GFP-labelled protein continues to move slowly with microtubule polymerization, while displaying visible diffusion near the tip (see Fig. 8c for the respective kymograph). This sequence was recorded using only a 488 nm laser, which excites both green and red fluorescence, subsequently separated using different emission. The coverslip-attached microtubule seed is seen as brighter red, because of a higher proportion of Rhodamine-labeled tubulin. Video is played 2x faster than recorded. Scale bar, 2 μm.
A complex of 3–4 FL CENP-E dimers tracks a dynamic microtubule tip for several phases of microtubule growth and shortening. Motions were recorded as in Supplementary Video S6, and played 20x faster than recorded. Scale bar, 2 μm.
Motor-coated bead (0.5 μm) walks to the plus end of a microtubule, which was grown from a coverslip-attached axoneme and stabilized temporarily with a GMPCPP-containing, Rhodamine-labelled cap (arrowhead). With bright illumination, the Rhodamine-labelled microtubule cap is destroyed (images in red), triggering microtubule depolymerization and subsequent bead detachment. Time-lapse images were acquired with low-light DIC, so individual microtubules are hard to discern. Video was recorded at 1 fps with 60 millisec exposure, played 7.5x faster than recorded.
Processive motion of the bead coated with full-length CENP-E towards the microtubule plus and then minus end.
This video shows an analogous experiment to that in Supplementary Video S8 but with the bead coated with FL CENP-E. The bead is seen moving on the microtubule wall towards the microtubule plus end, but when the microtubule begins to shorten, the bead moves in the minus-end direction in conjunction with microtubule disassembly. Arrow marks the initial position of the bead. Played 15x faster than recorded.
This video shows maximal projections of 3 confocal slices (0.5 μm per step) of HeLa cells expressing Mis12–GFP (kinetochore marker). Cells were treated with monastrol to create the mitotic configuration in which chromosomes are gathered in a rosette around the unseparated poles. Microtubule depolymerization was induced by nocodazole at the beginning of these two sequences and motions of the kinetochores were recorded every 10 sec for 5 min. The inward chromosome motion is greatly reduced in the cell treated with CENP-E inhibitor. Video is played 150x faster than recorded and repeated 3 times. Scale bar, 5 μm.
Motor domains are shown in blue and orange, CENP-E tails are pink and the coiled coil is grey. To simplify calculations, the motor walks along the same protofilament as the diffusing tail. Fluctuations in the length of microtubule protofilaments reflect fast tubulin exchange at the microtubule tip. This sequence shows the molecular events at a millisecond timescale, when microtubule elongation and motor walking appear much slower relative to the tubulin exchange and the diffusion and binding/unbinding of the tail. After reaching the end of the protofilament track, the motor domains dissociate rapidly (with the rate of tubulin exchange) but remain tethered to the microtubule lattice via the diffusing tail. The tail’s residency time on the microtubule lattice is 0.5 sec, whereas the motor re-binding to the lattice takes only 1 millisec, so the motor is highly likely to resume the plus-end-directed motion before the tail loses its microtubule association. Calculations were carried out with parameter values specified in Table S3 and with a 20 μsec time step; only 1 out of every 50 frames is shown. Video is played at 10 fps, so 1 sec of the video corresponds to 10 millisec of the ‘model’ time.
This simulation is for a longer time than in Supplementary Video S11, and it is played ‘faster’ to show the relatively slow processes of microtubule elongation, shortening and CENP-E tip-tracking. For simplicity, the elongating and shortening microtubule ends were modeled similarly, but with different constants of tubulin exchange. The CENP-E motor repeatedly walks to the end of the protofilament and detaches, but it remains tethered and quickly resumes its walking. This ‘tethered motor’ tip-tracking mechanism requires both the motor and tail CENP-E domains, but it does not require that the CENP-E can discriminate between the microtubule wall and its growing or shortening ends. The video was generated as Supplementary Video S11, but only 1 out of every 2,500 frames is shown at 10 fps.
About this article
Nature Chemical Biology (2017)