Mechanistic principles regulating microtubule network organization in metaphase and anaphase

Abstract

The proper organization of the microtubule-based spindle during cell division requires the collective activity of many different proteins. These include motor and non-motor microtubule-associated proteins (MAPs) whose functions include sliding and crosslinking microtubules to assemble microtubule arrays and regulate filament sliding rates. However, the mechanisms of how these proteins function cooperatively to control the timing of mitotic processes such as chromosome segregation and spindle elongation remain unclear. While motors such as Eg5 actively generate outward forces in the spindle, passive crosslinkers such as PRC1 must form stable midzone overlaps and resist microtubule sliding. PRC1 is a MAP that preferentially crosslinks overlapping antiparallel microtubules at the spindle midzone. It has been proposed to act as a molecular brake in anaphase, but insight into how it does so is lacking. To address these gaps in knowledge, we first employed a modified microtubule gliding assay to rupture PRC1-mediated microtubule pairs using surface-bound kinesins. We discovered that PRC1 crosslinks always reduce bundled filament sliding velocities relative to single microtubule gliding rates and do so via two distinct emergent modes of mechanical resistance to motor-driven sliding. We term these behaviors braking and coasting, where braking events exhibit substantially slowed microtubule sliding compared to coasting events. Strikingly, braking behavior requires the formation of two distinct high-density clusters of PRC1 molecules near microtubule tips. Our results suggest a cooperative mechanism for PRC1 accumulation when under mechanical load that leads to a unique state of enhanced resistance to filament sliding and provides insight into collective protein ensemble behavior in regulating the mechanics of spindle assembly. We expanded on these finding by next investigating PRC1’s biochemical regulation and its impact on PRC1’s spatiotemporal localization and function in the spindle. In metaphase, PRC1 is phosphorylated by cyclin-dependent kinase and is found in bridging fibers, structures that laterally reinforce kinetochore fibers. To elucidate what changes phosphorylation induces in PRC1’s behavior, we used a phosphomimetic PRC1 construct in our rupture assays and also measured its bundling capacity. We found that while recruitment and retention of phosphomimetic PRC1 in sliding overlaps is greater than wildtype dephosphorylated PRC1, it exhibits weaker resistance to motor-driven microtubule sliding. Phosphomimetic PRC1 does form clusters at the overlap edges, and these clusters contain more molecules than in wildtype PRC1 bundles, but they do not contribute to its braking ability as in wildtype bundles. Wildtype PRC1 also bundles a wider variety of bundle architectures, while phosphomimetic PRC1 is more limited in the construction of its bundles. These results suggest that phosphorylation of PRC1 at its CDK sites may modulate its capacity to bundle microtubules and resist sliding, as it serves a more structural role in metaphase, and dephosphorylation may switch on its enhanced braking ability to regulate filament sliding in the midzone. To build on these results, we have also begun expressing these constructs in cells to better understand the mechanisms of PRC1’s function in the context of dynamic spindles. Other work presented here focuses on the mechanisms of force generation by PRC1 and Eg5. Using an optical trap to measure force output by PRC1 crosslinks in microtubule pairs, we observed that force production is independent of overlap length and PRC1 density, but does scale with increasing numbers of molecules and microtubule sliding velocity. Because PRC1’s resistive force is proportional to sliding velocity, it likely acts as a viscous dashpot in overlaps. We also observed that when sliding was paused, PRC1 molecules underwent rearrangement to produce greater forces upon resumption of sliding. This rearrangement may build higher-order structures that produce greater resistance, similar to the clusters observed in our rupture assays. PRC1’s dashpot-like behavior may enable it to adapt to a variety of motor proteins in the spindle with different stepping velocities. As mentioned, Eg5 is one motor present in the spindle that is involved in maintenance of spindle length and bipolarity in metaphase and microtubule sliding in anaphase. Eg5 is well-characterized but the exact function of its tail domain and the mechanism of its conserved microtubule sliding ability are poorly understood. We employed optical trapping to measure the force production of full length Eg5 and a construct missing the tail domain. First, we found that 50-100 fold more of the tailless construct was required to achieve the same sliding as full length Eg5. However, similar amounts of motors were found in overlaps of similar lengths, with longer overlaps resulting in larger force plateaus. Tailless Eg5 also increased its force output in shorter bursts, ultimately reaching lower plateaus than the full length construct. This suggests that Eg5’s tail is important for microtubule sliding and aids in the regulation of force production during sliding. Tailless motors are unable to engage both microtubules in a pair to slide them relative to one another; coupled with an increase in directional switching, this renders tailless Eg5 unable to produce substantial sliding forces in microtubule overlaps. These results were part of a larger collaborative project; together, this work suggested a new mechanism for kinesin-5 regulation that occurs via its tail domain, which binds the motor domain to stabilize it when bound to a microtubule. This slows motility for greater crosslinking, and consequently greater capacity to generate force to slide microtubules apart. All together, the research presented here employs a combination of in vitro, in vivo, biophysical, and biochemical approaches to probe the mechanisms governing the behavior of well-known and essential proteins in the spindle.