Coordinated poleward flux of sister kinetochore fibers drives chromosome alignment

Abstract

Chromosome alignment at the spindle equator during metaphase is the most remarkable feature of mitosis, which promotes proper chromosome segregation and depends on the forces exerted at the plus end of kinetochore microtubules and polar ejection forces. However, forces arising from lateral mechanical coupling of kinetochore fibers with non-kinetochore microtubules play a role in chromosome alignment, but the mechanism is unclear. Here we develop a speckle microscopy assay to measure the poleward flux of individual microtubules in spindles of human cells and show that bridging microtubules slide apart and undergo poleward flux at a higher rate than kinetochore microtubules. Depletion of the microtubule coupler NuMa increased the difference in the flux velocity of kinetochore and bridging microtubules, suggesting that sliding forces from the bridging fiber are transmitted largely through NuMa onto the associated kinetochore fibers. Depletions of Kif18A/kinesin-8, Kif4A/kinesin-4, as well as double depletions of Kif18A together with Kif4A or Kif18A together with the crosslinker of antiparallel microtubules PRC1 increased the flux velocity of kinetochore fibers up to the velocity of bridging fibers, due to the increased coupling resulting from the extended antiparallel overlap regions. We found severe kinetochore misalignment after double Kif18A and Kif4A as well as Kif18A and PRC1 depletions compared to a single Kif18A depletion, suggesting that forces within the bridging fiber have a centering effect on the kinetochores. We propose that lateral length-dependent sliding forces that the bridging fiber exerts onto kinetochore fibers drive the movement of kinetochores towards the spindle center, thereby promoting chromosome alignment.

Biomechanics of chromosome alignment at the spindle midplane

“…This leads to a higher poleward flux velocity of the longer k-fiber and directs the net force towards the spindle equator. This was corroborated by speckle microscopy experiments that revealed a difference in the poleward flux of longer and shorter k-fibers in human spindles 35 .…”

“…The total force on the chromosome is then directed towards the spindle center and proportional to the difference in the length of the overlap on either side (Figure 2B). Such a centering mechanism was recently proposed for spindles in human cells 35 .…”

“…Bridging microtubules are mainly nucleated in an augmindependent manner 100 , in agreement with the localization of their minus ends along kinetochore microtubules. Crosslinking of parallel overlap regions between bridging and kinetochore microtubules is mediated by NuMa 35,101 , whereas anti-parallel overlaps within bridging fibers are linked together by the protein regulator of cytokinesis 1 (PRC1). Several motor proteins that slide microtubules or regulate microtubule dynamics are also found within the bridging fiber (Figure 3D), including Eg5 33,102 , CENP-E 82 , Kif4A, Kif18A, and MKLP1 103 .…”

“…Eg5 is likely the main microtubule slider as its inactivation during metaphase results in spindle shortening and collapse 104 . The reduced poleward flux velocity of the bridging microtubules observed after CENP-E depletion suggests that this motor also has a role in sliding of bridging microtubules 35 . The overlap length of the antiparallel overlaps within the bridging fiber is regulated by Kif4A and Kif18A 103 .…”

“…As PRC1-labeled bridging fibers show one-to-one association with a pair of sister k-fibers 106 , this could give rise to flux-dependent equalization of tension at kinetochores 13 and a closed-loop force network independent of centrosomes 107 . Indeed, it was shown that bridging microtubules slide apart and serve as a platform for force generation that underlies microtubule poleward flux 35 . Upon loss of k-fibers in the spindle, bridging fibers undergo similar rates of poleward flux, suggesting that bridging fiber flux is independent of k-fibers.…”